67 results on '"trajectory forecasting"'
Search Results
2. Toward Smart Doors: A Position Paper
- Author
-
Capogrosso, Luigi, Skenderi, Geri, Girella, Federico, Fummi, Franco, Cristani, Marco, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Rousseau, Jean-Jacques, editor, and Kapralos, Bill, editor
- Published
- 2023
- Full Text
- View/download PDF
3. Transformer Networks for Future Person Localization in First-Person Videos
- Author
-
Alikadic, Amar, Saito, Hideo, Hachiuma, Ryo, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Bebis, George, editor, Li, Bo, editor, Yao, Angela, editor, Liu, Yang, editor, Duan, Ye, editor, Lau, Manfred, editor, Khadka, Rajiv, editor, Crisan, Ana, editor, and Chang, Remco, editor
- Published
- 2022
- Full Text
- View/download PDF
4. Aware of the History: Trajectory Forecasting with the Local Behavior Data
- Author
-
Zhong, Yiqi, Ni, Zhenyang, Chen, Siheng, Neumann, Ulrich, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Avidan, Shai, editor, Brostow, Gabriel, editor, Cissé, Moustapha, editor, Farinella, Giovanni Maria, editor, and Hassner, Tal, editor
- Published
- 2022
- Full Text
- View/download PDF
5. Social-Implicit: Rethinking Trajectory Prediction Evaluation and The Effectiveness of Implicit Maximum Likelihood Estimation
- Author
-
Mohamed, Abduallah, Zhu, Deyao, Vu, Warren, Elhoseiny, Mohamed, Claudel, Christian, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Avidan, Shai, editor, Brostow, Gabriel, editor, Cissé, Moustapha, editor, Farinella, Giovanni Maria, editor, and Hassner, Tal, editor
- Published
- 2022
- Full Text
- View/download PDF
6. PreTraM: Self-supervised Pre-training via Connecting Trajectory and Map
- Author
-
Xu, Chenfeng, Li, Tian, Tang, Chen, Sun, Lingfeng, Keutzer, Kurt, Tomizuka, Masayoshi, Fathi, Alireza, Zhan, Wei, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Avidan, Shai, editor, Brostow, Gabriel, editor, Cissé, Moustapha, editor, Farinella, Giovanni Maria, editor, and Hassner, Tal, editor
- Published
- 2022
- Full Text
- View/download PDF
7. ST‐SIGMA: Spatio‐temporal semantics and interaction graph aggregation for multi‐agent perception and trajectory forecasting
- Author
-
Yang Fang, Bei Luo, Ting Zhao, Dong He, Bingbing Jiang, and Qilie Liu
- Subjects
feature fusion ,graph interaction ,hierarchical aggregation ,scene perception ,scene semantics ,trajectory forecasting ,Computational linguistics. Natural language processing ,P98-98.5 ,Computer software ,QA76.75-76.765 - Abstract
Abstract Scene perception and trajectory forecasting are two fundamental challenges that are crucial to a safe and reliable autonomous driving (AD) system. However, most proposed methods aim at addressing one of the two challenges mentioned above with a single model. To tackle this dilemma, this paper proposes spatio‐temporal semantics and interaction graph aggregation for multi‐agent perception and trajectory forecasting (ST‐SIGMA), an efficient end‐to‐end method to jointly and accurately perceive the AD environment and forecast the trajectories of the surrounding traffic agents within a unified framework. ST‐SIGMA adopts a trident encoder–decoder architecture to learn scene semantics and agent interaction information on bird’s‐eye view (BEV) maps simultaneously. Specifically, an iterative aggregation network is first employed as the scene semantic encoder (SSE) to learn diverse scene information. To preserve dynamic interactions of traffic agents, ST‐SIGMA further exploits a spatio‐temporal graph network as the graph interaction encoder. Meanwhile, a simple yet efficient feature fusion method to fuse semantic and interaction features into a unified feature space as the input to a novel hierarchical aggregation decoder for downstream prediction tasks is designed. Extensive experiments on the nuScenes data set have demonstrated that the proposed ST‐SIGMA achieves significant improvements compared to the state‐of‐the‐art (SOTA) methods in terms of scene perception and trajectory forecasting, respectively. Therefore, the proposed approach outperforms SOTA in terms of model generalisation and robustness and is therefore more feasible for deployment in real‐world AD scenarios.
- Published
- 2022
- Full Text
- View/download PDF
8. ST‐SIGMA: Spatio‐temporal semantics and interaction graph aggregation for multi‐agent perception and trajectory forecasting.
- Author
-
Fang, Yang, Luo, Bei, Zhao, Ting, He, Dong, Jiang, Bingbing, and Liu, Qilie
- Subjects
SEMANTICS ,AUTONOMOUS vehicles ,FORECASTING ,GENERALIZATION ,CHANNEL coding - Abstract
Scene perception and trajectory forecasting are two fundamental challenges that are crucial to a safe and reliable autonomous driving (AD) system. However, most proposed methods aim at addressing one of the two challenges mentioned above with a single model. To tackle this dilemma, this paper proposes spatio‐temporal semantics and interaction graph aggregation for multi‐agent perception and trajectory forecasting (ST‐SIGMA), an efficient end‐to‐end method to jointly and accurately perceive the AD environment and forecast the trajectories of the surrounding traffic agents within a unified framework. ST‐SIGMA adopts a trident encoder–decoder architecture to learn scene semantics and agent interaction information on bird's‐eye view (BEV) maps simultaneously. Specifically, an iterative aggregation network is first employed as the scene semantic encoder (SSE) to learn diverse scene information. To preserve dynamic interactions of traffic agents, ST‐SIGMA further exploits a spatio‐temporal graph network as the graph interaction encoder. Meanwhile, a simple yet efficient feature fusion method to fuse semantic and interaction features into a unified feature space as the input to a novel hierarchical aggregation decoder for downstream prediction tasks is designed. Extensive experiments on the nuScenes data set have demonstrated that the proposed ST‐SIGMA achieves significant improvements compared to the state‐of‐the‐art (SOTA) methods in terms of scene perception and trajectory forecasting, respectively. Therefore, the proposed approach outperforms SOTA in terms of model generalisation and robustness and is therefore more feasible for deployment in real‐world AD scenarios. [ABSTRACT FROM AUTHOR]
- Published
- 2022
- Full Text
- View/download PDF
9. Multi-Camera Trajectory Forecasting With Trajectory Tensors.
- Author
-
Styles, Olly, Guha, Tanaya, and Sanchez, Victor
- Subjects
- *
TRAFFIC monitoring , *FORECASTING , *VIDEO compression , *CAMERAS , *VIDEO coding , *DIGITAL cameras - Abstract
We introduce the problem of multi-camera trajectory forecasting (MCTF), which involves predicting the trajectory of a moving object across a network of cameras. While multi-camera setups are widespread for applications such as surveillance and traffic monitoring, existing trajectory forecasting methods typically focus on single-camera trajectory forecasting (SCTF), limiting their use for such applications. Furthermore, using a single camera limits the field-of-view available, making long-term trajectory forecasting impossible. We address these shortcomings of SCTF by developing an MCTF framework that simultaneously uses all estimated relative object locations from several viewpoints and predicts the object's future location in all possible viewpoints. Our framework follows a Which-When-Where approach that predicts in which camera(s) the objects appear and when and where within the camera views they appear. To this end, we propose the concept of trajectory tensors: a new technique to encode trajectories across multiple camera views and the associated uncertainties. We develop several encoder-decoder MCTF models for trajectory tensors and present extensive experiments on our own database (comprising 600 hours of video data from 15 camera views) created particularly for the MCTF task. Results show that our trajectory tensor models outperform coordinate trajectory-based MCTF models and existing SCTF methods adapted for MCTF. Code is available from: https://github.com/olly-styles/Trajectory-Tensors. [ABSTRACT FROM AUTHOR]
- Published
- 2022
- Full Text
- View/download PDF
10. Taming the Long Tail of Deep Probabilistic Forecasting
- Author
-
Sharan, Mayank
- Subjects
Computer science ,Deep Probabilistic Forecasting ,Long Tail ,Time Series Forecasting ,Trajectory Forecasting - Abstract
Deep probabilistic forecasting is gaining attention across numerous applications. From weather prognosis, electricity consumption estimation, traffic flow prediction, to autonomous vehicle trajectory prediction. However, existing approaches focus on improving on average metrics without addressing the long-tailed distribution of errors. This thesis identifies long tail behavior in the error distribution of state-of-the-art deep learning methods for probabilistic forecasting. We analyze potential sources and explanations for this behavior. Further, we present two loss augmentation methods to mitigate tailedness pf error distributions: Pareto Loss and Kurtosis Loss. Both methods modify the loss using the concept of moments to penalize higher loss samples. Kurtosis Loss uses a symmetric measure, the fourth moment, while Pareto Loss uses an asymmetric measure of right-tailedness and models loss using a Generalized Pareto Distribution (GPD). We demonstrate the performance of our methods on several real-world datasets, including time series and spatiotemporal trajectories, achieving significant improvements on tail error metrics, while maintaining and even improving upon average error metrics.
- Published
- 2023
11. OpenTraj: Assessing Prediction Complexity in Human Trajectories Datasets
- Author
-
Amirian, Javad, Zhang, Bingqing, Castro, Francisco Valente, Baldelomar, Juan José, Hayet, Jean-Bernard, Pettré, Julien, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Woeginger, Gerhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Ishikawa, Hiroshi, editor, Liu, Cheng-Lin, editor, Pajdla, Tomas, editor, and Shi, Jianbo, editor
- Published
- 2021
- Full Text
- View/download PDF
12. Human Trajectory Forecasting in Crowds: A Deep Learning Perspective.
- Author
-
Kothari, Parth, Kreiss, Sven, and Alahi, Alexandre
- Abstract
Since the past few decades, human trajectory forecasting has been a field of active research owing to its numerous real-world applications: evacuation situation analysis, deployment of intelligent transport systems, traffic operations, to name a few. In this work, we cast the problem of human trajectory forecasting as learning a representation of human social interactions. Early works handcrafted this representation based on domain knowledge. However, social interactions in crowded environments are not only diverse but often subtle. Recently, deep learning methods have outperformed their handcrafted counterparts, as they learn about human-human interactions in a more generic data-driven fashion. In this work, we present an in-depth analysis of existing deep learning-based methods for modelling social interactions. We propose two domain-knowledge inspired data-driven methods to effectively capture these social interactions. To objectively compare the performance of these interaction-based forecasting models, we develop a large scale interaction-centric benchmark TrajNet++, a significant yet missing component in the field of human trajectory forecasting. We propose novel performance metrics that evaluate the ability of a model to output socially acceptable trajectories. Experiments on TrajNet++ validate the need for our proposed metrics, and our method outperforms competitive baselines on both real-world and synthetic datasets. [ABSTRACT FROM AUTHOR]
- Published
- 2022
- Full Text
- View/download PDF
13. An Adversarial Learned Trajectory Predictor with Knowledge-Rich Latent Variables
- Author
-
He, Caizhen, Yang, Biao, Chen, Lanping, Yan, Guocheng, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Woeginger, Gerhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Peng, Yuxin, editor, Liu, Qingshan, editor, Lu, Huchuan, editor, Sun, Zhenan, editor, Liu, Chenglin, editor, Chen, Xilin, editor, Zha, Hongbin, editor, and Yang, Jian, editor
- Published
- 2020
- Full Text
- View/download PDF
14. Diverse and Admissible Trajectory Forecasting Through Multimodal Context Understanding
- Author
-
Park, Seong Hyeon, Lee, Gyubok, Seo, Jimin, Bhat, Manoj, Kang, Minseok, Francis, Jonathan, Jadhav, Ashwin, Liang, Paul Pu, Morency, Louis-Philippe, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Woeginger, Gerhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Vedaldi, Andrea, editor, Bischof, Horst, editor, Brox, Thomas, editor, and Frahm, Jan-Michael, editor
- Published
- 2020
- Full Text
- View/download PDF
15. Dynamic and Static Context-Aware LSTM for Multi-agent Motion Prediction
- Author
-
Tao, Chaofan, Jiang, Qinhong, Duan, Lixin, Luo, Ping, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Woeginger, Gerhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Vedaldi, Andrea, editor, Bischof, Horst, editor, Brox, Thomas, editor, and Frahm, Jan-Michael, editor
- Published
- 2020
- Full Text
- View/download PDF
16. Trajectron++: Dynamically-Feasible Trajectory Forecasting with Heterogeneous Data
- Author
-
Salzmann, Tim, Ivanovic, Boris, Chakravarty, Punarjay, Pavone, Marco, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Woeginger, Gerhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Vedaldi, Andrea, editor, Bischof, Horst, editor, Brox, Thomas, editor, and Frahm, Jan-Michael, editor
- Published
- 2020
- Full Text
- View/download PDF
17. An RNN-Based IMM Filter Surrogate
- Author
-
Becker, Stefan, Hug, Ronny, Hübner, Wolfgang, Arens, Michael, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Woeginger, Gerhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Felsberg, Michael, editor, Forssén, Per-Erik, editor, Sintorn, Ida-Maria, editor, and Unger, Jonas, editor
- Published
- 2019
- Full Text
- View/download PDF
18. RED: A Simple but Effective Baseline Predictor for the TrajNet Benchmark
- Author
-
Becker, Stefan, Hug, Ronny, Hübner, Wolfgang, Arens, Michael, Hutchison, David, Series Editor, Kanade, Takeo, Series Editor, Kittler, Josef, Series Editor, Kleinberg, Jon M., Series Editor, Mattern, Friedemann, Series Editor, Mitchell, John C., Series Editor, Naor, Moni, Series Editor, Pandu Rangan, C., Series Editor, Steffen, Bernhard, Series Editor, Terzopoulos, Demetri, Series Editor, Tygar, Doug, Series Editor, Leal-Taixé, Laura, editor, and Roth, Stefan, editor
- Published
- 2019
- Full Text
- View/download PDF
19. A Novel Trajectory Generator Based on a Constrained GAN and a Latent Variables Predictor
- Author
-
Wei Wu, Biao Yang, Dong Wang, and Weigong Zhang
- Subjects
Trajectory forecasting ,generative adversarial network ,latent variable predictor ,future uncertainty ,Electrical engineering. Electronics. Nuclear engineering ,TK1-9971 - Abstract
Forecasting pedestrian trajectory is critical for versatile applications, such as autonomous driving and social robot, when they work in human-centric environments. However, it is challenging to predict pedestrians' future trajectories due to the inherent human properties and pedestrians' social interactions. Recent works predict future trajectories by using a generative model, which captures social interactions with pooling- or graph-based strategies and generates multi-modal outputs with latent variables sampled from random Gaussian noise. Nevertheless, they introduce little human knowledge, which is beneficial for improved prediction performance. In this work, we propose to learn informative latent variables from pedestrians' future trajectories. Moreover, we present a distance-direction pooling module, which captures social interactions in a more intuitive manner. Besides, we introduce an additional constraint on generative adversarial network optimization to generate more realistic results. Two benchmarking datasets, ETH (Pellegrini et al., 2010) and UCY (Leal-Taixé et al., 2014), are used to evaluate the proposed method. Comparisons between our method and several state-of-the-art methods demonstrate the superiority of the proposed method in generating more accurate future trajectories.
- Published
- 2020
- Full Text
- View/download PDF
20. r2p2: A ReparameteRized Pushforward Policy for Diverse, Precise Generative Path Forecasting
- Author
-
Rhinehart, Nicholas, Kitani, Kris M., Vernaza, Paul, Hutchison, David, Series Editor, Kanade, Takeo, Series Editor, Kittler, Josef, Series Editor, Kleinberg, Jon M., Series Editor, Mattern, Friedemann, Series Editor, Mitchell, John C., Series Editor, Naor, Moni, Series Editor, Pandu Rangan, C., Series Editor, Steffen, Bernhard, Series Editor, Terzopoulos, Demetri, Series Editor, Tygar, Doug, Series Editor, Weikum, Gerhard, Series Editor, Ferrari, Vittorio, editor, Hebert, Martial, editor, Sminchisescu, Cristian, editor, and Weiss, Yair, editor
- Published
- 2018
- Full Text
- View/download PDF
21. Forecasting People Trajectories and Head Poses by Jointly Reasoning on Tracklets and Vislets.
- Author
-
Hasan, Irtiza, Setti, Francesco, Tsesmelis, Theodore, Belagiannis, Vasileios, Amin, Sikandar, Del Bue, Alessio, Cristani, Marco, and Galasso, Fabio
- Subjects
- *
FORECASTING , *COVARIANCE matrices , *CAPABILITIES approach (Social sciences) , *TIME perspective , *SOCIAL interaction , *PEDESTRIANS - Abstract
In this article, we explore the correlation between people trajectories and their head orientations. We argue that people trajectory and head pose forecasting can be modelled as a joint problem. Recent approaches on trajectory forecasting leverage short-term trajectories (aka tracklets) of pedestrians to predict their future paths. In addition, sociological cues, such as expected destination or pedestrian interaction, are often combined with tracklets. In this article, we propose MiXing-LSTM (MX-LSTM) to capture the interplay between positions and head orientations (vislets) thanks to a joint unconstrained optimization of full covariance matrices during the LSTM backpropagation. We additionally exploit the head orientations as a proxy for the visual attention, when modeling social interactions. MX-LSTM predicts future pedestrians location and head pose, increasing the standard capabilities of the current approaches on long-term trajectory forecasting. Compared to the state-of-the-art, our approach shows better performances on an extensive set of public benchmarks. MX-LSTM is particularly effective when people move slowly, i.e., the most challenging scenario for all other models. The proposed approach also allows for accurate predictions on a longer time horizon. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF
22. PHEV Power Management Optimization Using Trajectory Forecasting Based Machine Learning
- Author
-
Garcia, Joseph Augusto
- Subjects
Mechanical engineering ,CNN ,Machine Learning ,PHEV ,Trajectory Forecasting - Abstract
In hopes of lessening the reliance on fossil fuels, Plug-in Hybrid Electric Vehicles (PHEVs) have become an attractive option as an alternative fuel vehicle due to their larger electric motors and energy storage systems (ESS). PHEVs can propel themself relying solely on their internal combustion engine (ICE), electric motor (EM), and or a hybrid of both. To improve their fuel efficiency, many studies have been done to investigate the use of a priori route information to optimize the use of a PHEV’s ICE and EM. This study introduces a real-time machine learning application of a control strategy known as Trajectory Forecasting (TF). TF takes a priori knowledge of a PHEV’s pre-planned route to determine when the vehicle will use its different forms of propulsion in the form of propulsion mode scheduling. However, it assumes constant route data such as traffic and resulting driving speed for its scheduling to be applicable. To automatically account for changing traffic as well as choose better alternative routes, this study looks at the use of a Convolutional Neural Network (CNN) to simulate a PHEV’s operation along available routes beforehand according to the rules of TF to choose a route that best satisfies a driver’s want, better fuel efficiency and possibly lower emissions. This new real-time TF-based machine learning control strategy is evaluated and compared to common PHEV control strategies such as Charge Sustaining (CS) and Charge Depletion (CD) using National Renewable Energy Laboratory’s vehicle simulator ADVISOR. Results show possible increases in Mpgge from 1.72%-130%, decreases in emitted hydrocarbons (HC), carbon monoxide (CO), and nitrous oxides (NOx) from 0.05%-70%, and 0.05%-90.35% reduction in gasoline consumption depending on overall route length and PHEV configuration.
- Published
- 2021
23. End-to-End Pedestrian Trajectory Forecasting with Transformer Network
- Author
-
Hai-Yan Yao, Wang-Gen Wan, and Xiang Li
- Subjects
trajectory forecasting ,transformer ,random deviation query ,Geography (General) ,G1-922 - Abstract
Analysis of pedestrians’ motion is important to real-world applications in public scenes. Due to the complex temporal and spatial factors, trajectory prediction is a challenging task. With the development of attention mechanism recently, transformer network has been successfully applied in natural language processing, computer vision, and audio processing. We propose an end-to-end transformer network embedded with random deviation queries for pedestrian trajectory forecasting. The self-correcting scheme can enhance the robustness of the network. Moreover, we present a co-training strategy to improve the training effect. The whole scheme is trained collaboratively by the original loss and classification loss. Therefore, we also achieve more accurate prediction results. Experimental results on several datasets indicate the validity and robustness of the network. We achieve the best performance in individual forecasting and comparable results in social forecasting. Encouragingly, our approach achieves a new state of the art on the Hotel and Zara2 datasets compared with the social-based and individual-based approaches.
- Published
- 2022
- Full Text
- View/download PDF
24. Trajectory Forecasting With Neural Networks: An Empirical Evaluation and A New Hybrid Model.
- Author
-
Wang, Yuan, Zhang, Dongxiang, Liu, Ying, and Tan, Kian-Lee
- Abstract
Recent years have witnessed the advent and prevalence of deep learning that have provoked storms in ITS (Intelligent Transportation Systems). Consequently, traditional ML models in many applications have been replaced by the new learning techniques. In this paper, we focus on the problem of trajectory prediction, which is a cornerstone component to support many useful higher-level applications in ITS. We conduct so far the most comprehensive evaluation on various models proposed for trajectory prediction, including 5 statistic-based models and 10 deep-learning based networks. In addition, we propose a hybrid model that integrates MLP to extract local features and LSTM to capture long term dependency. It is also enhanced with a moving-state prediction model based on random forest to further reduce prediction error. We conduct extensive empirical evaluation on three real datasets with different motion patterns and reveal interesting findings. Furthermore, the experimental results show that our proposed hybrid model achieves superior performance over its competitors in all the datasets. [ABSTRACT FROM AUTHOR]
- Published
- 2020
- Full Text
- View/download PDF
25. Learning Social Etiquette: Human Trajectory Understanding In Crowded Scenes
- Author
-
Robicquet, Alexandre, Sadeghian, Amir, Alahi, Alexandre, Savarese, Silvio, Hutchison, David, Series editor, Kanade, Takeo, Series editor, Kittler, Josef, Series editor, Kleinberg, Jon M., Series editor, Mattern, Friedemann, Series editor, Mitchell, John C., Series editor, Naor, Moni, Series editor, Pandu Rangan, C., Series editor, Steffen, Bernhard, Series editor, Terzopoulos, Demetri, Series editor, Tygar, Doug, Series editor, Weikum, Gerhard, Series editor, Leibe, Bastian, editor, Matas, Jiri, editor, Sebe, Nicu, editor, and Welling, Max, editor
- Published
- 2016
- Full Text
- View/download PDF
26. Real-time forecasting of driver-vehicle dynamics on 3D roads: A deep-learning framework leveraging Bayesian optimisation.
- Author
-
Paparusso, Luca, Melzi, Stefano, and Braghin, Francesco
- Subjects
- *
AUTOMOBILE driving simulators , *DEEP learning , *FORECASTING , *SYSTEM dynamics , *RECURRENT neural networks - Abstract
Most state-of-the-art works in trajectory forecasting for automotive target predicting the pose and orientation of the agents in the scene. This represents a particularly useful problem, for instance in autonomous driving, but it does not cover a spectrum of applications in control and simulation that require information on vehicle dynamics features other than pose and orientation. Also, multi-step dynamic simulation of complex multibody models does not seem to be a viable solution for real-time long-term prediction, due to the high computational time required. To bridge this gap, we present a deep-learning framework to model and predict the evolution of the coupled driver-vehicle system dynamics jointly on a complex road geometry. It consists of two components. The first, a neural network predictor, is based on Long Short-Term Memory autoencoders and fuses the information on the road geometry and the past driver-vehicle system dynamics to produce context-aware predictions. The second, a Bayesian optimiser, is proposed to tune some significant hyperparameters of the network. These govern the network complexity, as well as the features importance. The result is a self-tunable framework with real-time applicability, which allows the user to specify the features of interest. The approach has been validated with a case study centred on motion cueing algorithms, using a dataset collected during test sessions of a non-professional driver on a dynamic driving simulator. A 3D track with complex geometry has been employed as driving environment to render the prediction task challenging. Finally, the robustness of the neural network to changes in the driver and track was investigated to set guidelines for future works. • Deep learning can be used to forecast complex driver-vehicle dynamics jointly. • 3D road geometries can be encoded using Long Short-Term Memory autoencoders. • Neural networks can be faster than multibody models in long-term vehicle simulation. • Bayesian optimisation can balance model complexity for real-time applicability. [ABSTRACT FROM AUTHOR]
- Published
- 2023
- Full Text
- View/download PDF
27. Human Trajectory Forecasting in Crowds: A Deep Learning Perspective
- Author
-
Parth Kothari, Alexandre Alahi, and Sven Kreiss
- Subjects
FOS: Computer and information sciences ,Computer science ,business.industry ,Computer Vision and Pattern Recognition (cs.CV) ,Mechanical Engineering ,Deep learning ,Computer Science - Computer Vision and Pattern Recognition ,deep learning ,Machine learning ,computer.software_genre ,Field (computer science) ,Computer Science Applications ,benchmark ,Crowds ,Automotive Engineering ,Trajectory ,Benchmark (computing) ,social interactions ,Domain knowledge ,Artificial intelligence ,trajectory forecasting ,business ,Intelligent transportation system ,computer ,Situation analysis - Abstract
Since the past few decades, human trajectory forecasting has been a field of active research owing to its numerous real-world applications: evacuation situation analysis, deployment of intelligent transport systems, traffic operations, to name a few. Early works handcrafted this representation based on domain knowledge. However, social interactions in crowded environments are not only diverse but often subtle. Recently, deep learning methods have outperformed their handcrafted counterparts, as they learned about human-human interactions in a more generic data-driven fashion. In this work, we present an in-depth analysis of existing deep learning-based methods for modelling social interactions. We propose two knowledge-based data-driven methods to effectively capture these social interactions. To objectively compare the performance of these interaction-based forecasting models, we develop a large scale interaction-centric benchmark TrajNet++, a significant yet missing component in the field of human trajectory forecasting. We propose novel performance metrics that evaluate the ability of a model to output socially acceptable trajectories. Experiments on TrajNet++ validate the need for our proposed metrics, and our method outperforms competitive baselines on both real-world and synthetic datasets., IEEE format, Layer-wise Relevance Propagation added
- Published
- 2022
28. Deep Learning Approaches for Time-Evolving Scenarios
- Author
-
Bertugli, Alessia
- Subjects
Trajectory forecasting, continual learning, meta-learning, finance ,Trajectory forecasting ,meta-learning ,finance ,continual learning - Published
- 2023
- Full Text
- View/download PDF
29. CLIMAT:clinically-inspired multi-agent transformers for knee osteoarthritis trajectory forecasting
- Abstract
In medical applications, deep learning methods are designed to automate diagnostic tasks. However, a clinically relevant question that practitioners usually face, is how to predict the future trajectory of a disease (prognosis). Current methods for such a problem often require domain knowledge, and are complicated to apply. In this paper, we formulate the prognosis prediction problem as a one-to-many forecasting problem from multimodal data. Inspired by a clinical decision-making process with two agents — a radiologist and a general practitioner, we model a prognosis prediction problem with two transformer-based components that share information between each other. The first block in this model aims to analyze the imaging data, and the second block leverages the internal representations of the first one as inputs, also fusing them with auxiliary patient data. We show the effectiveness of our method in predicting the development of structural knee osteoarthritis changes over time. Our results show that the proposed method outperforms the state-of-the-art baselines in terms of various performance metrics. In addition, we empirically show that the existence of the multi-agent transformers with depths of 2 is sufficient to achieve good performances. Our code is publicly available at https://github.com/MIPT-Oulu/CLIMAT.
- Published
- 2022
30. Under the hood of transformer networks for trajectory forecasting.
- Author
-
Franco, Luca, Placidi, Leonardo, Giuliari, Francesco, Hasan, Irtiza, Cristani, Marco, and Galasso, Fabio
- Subjects
- *
CONVOLUTIONAL neural networks , *GENERATIVE adversarial networks , *SOURCE code , *SOCIAL interaction , *SOCIAL context - Abstract
• First in-depth experimental study of Transformer Networks and BERT for trajectory forecasting. • Comparative evaluation on the modelling of individual human trajectories. • First exhaustive evaluation of input/output representations and problem formulations. • Focus on the capability of Transformers to predict multi-modal futures. • Analysis of the impact of intention and long-term forecasting. Transformer Networks have established themselves as the de-facto state-of-the-art for trajectory forecasting but there is currently no systematic study on their capability to model the motion patterns of people, without interactions with other individuals nor the social context. There is abundant literature on LSTMs, CNNs and GANs on this subject. However methods adopting Transformer techniques achieve great performances by complex models and a clear analysis of their adoption as plain sequence models is missing. This paper proposes the first in-depth study of Transformer Networks (TF) and the Bidirectional Transformers (BERT) for the forecasting of the individual motion of people, without bells and whistles. We conduct an exhaustive evaluation of the input/output representations, problem formulations and sequence modelling, including a novel analysis of their capability to predict multi-modal futures. Out of comparative evaluation on the ETH+UCY benchmark, both TF and BERT are top performers in predicting individual motions and remain within a narrow margin wrt more complex techniques, including both social interactions and scene contexts. Source code will be released for all conducted experiments. [ABSTRACT FROM AUTHOR]
- Published
- 2023
- Full Text
- View/download PDF
31. CLIMAT:clinically-inspired multi-agent transformers for knee osteoarthritis trajectory forecasting
- Author
-
Nguyen, H. H. (Huy Hoang), Saarakkala, S. (Simo), Blaschko, M. B. (Matthew B.), and Tiulpin, A. (Aleksei)
- Subjects
osteoarthritis ,transformer ,deep learning ,prognosis ,trajectory forecasting - Abstract
In medical applications, deep learning methods are designed to automate diagnostic tasks. However, a clinically relevant question that practitioners usually face, is how to predict the future trajectory of a disease (prognosis). Current methods for such a problem often require domain knowledge, and are complicated to apply. In this paper, we formulate the prognosis prediction problem as a one-to-many forecasting problem from multimodal data. Inspired by a clinical decision-making process with two agents — a radiologist and a general practitioner, we model a prognosis prediction problem with two transformer-based components that share information between each other. The first block in this model aims to analyze the imaging data, and the second block leverages the internal representations of the first one as inputs, also fusing them with auxiliary patient data. We show the effectiveness of our method in predicting the development of structural knee osteoarthritis changes over time. Our results show that the proposed method outperforms the state-of-the-art baselines in terms of various performance metrics. In addition, we empirically show that the existence of the multi-agent transformers with depths of 2 is sufficient to achieve good performances. Our code is publicly available at https://github.com/MIPT-Oulu/CLIMAT.
- Published
- 2022
32. Asymmetrical Bi-RNN for pedestrian trajectory encoding
- Author
-
Rozenberg, Raphaël, Gesnouin, Joseph, Moutarde, Fabien, and Gesnouin, Joseph
- Subjects
[INFO.INFO-AI] Computer Science [cs]/Artificial Intelligence [cs.AI] ,FOS: Computer and information sciences ,I.2.10 ,Artificial Intelligence (cs.AI) ,Computer Science - Artificial Intelligence ,Computer Vision and Pattern Recognition (cs.CV) ,Pedestrian Safety ,Computer Science - Computer Vision and Pattern Recognition ,I.2.9 ,Sequence Encoding ,Trajectory Forecasting ,68T45 - Abstract
Pedestrian motion behavior involves a combination of individual goals and social interactions with other agents. In this article, we present an asymmetrical bidirectional recurrent neural network architecture called U-RNN to encode pedestrian trajectories and evaluate its relevance to replace LSTMs for various forecasting models. Experimental results on the Trajnet++ benchmark show that the U-LSTM variant yields better results regarding every available metrics (ADE, FDE, Collision rate) than common trajectory encoders for a variety of approaches and interaction modules, suggesting that the proposed approach is a viable alternative to the de facto sequence encoding RNNs. Our implementation of the asymmetrical Bi-RNNs for the Trajnet++ benchmark is available at: github.com/JosephGesnouin/Asymmetrical-Bi-RNNs-to-encode-pedestrian-trajectories, Comment: 7 pages
- Published
- 2021
- Full Text
- View/download PDF
33. Algorithm for the application of automatic driving technology and predicting the trajectory of movement
- Subjects
авÑомаÑиÑеÑкое вождение ,machine learning ,automatic driving ,компÑÑÑеÑное зÑение ,пÑогнозиÑование ÑÑаекÑоÑии Ð´Ð²Ð¸Ð¶ÐµÐ½Ð¸Ñ ,маÑинное обÑÑение ,trajectory forecasting ,computer vision - Abstract
ÐвалиÑикаÑÐ¸Ð¾Ð½Ð½Ð°Ñ ÑабоÑа на ÑемÑ: «ÐлгоÑиÑм пÑÐ¸Ð¼ÐµÐ½ÐµÐ½Ð¸Ñ ÑÐµÑ Ð½Ð¸ÐºÐ¸ Ð²Ð¾Ð¶Ð´ÐµÐ½Ð¸Ñ Ð¸ пÑогнозиÑÐ¾Ð²Ð°Ð½Ð¸Ñ ÑÑаекÑоÑии движениÑ». РнаÑÑном пÑоекÑе оÑÑажаеÑÑÑ Ð¿Ñименение алгоÑиÑмов компÑÑÑеÑной ÑиÑÑÐµÐ¼Ñ Ð·ÑÐµÐ½Ð¸Ñ Ð¸ подÑаздел иÑкÑÑÑÑвенного инÑеллекÑа â маÑинное обÑÑение. ÐÑ Ð¿ÑÐ¸Ð¼ÐµÐ½ÐµÐ½Ð¸Ñ Ð·Ð°ÐºÐ»ÑÑаеÑÑÑ Ð² ÑÑÑановлении ÑакÑоÑов Ñ ÑелÑÑ Ð¸ÑполÑÐ·Ð¾Ð²Ð°Ð½Ð¸Ñ Ð±ÐµÑпилоÑнÑÑ Ð°Ð²Ñомобилей или дÑÑÐ³Ð¸Ñ ÑÑанÑпоÑÑнÑÑ ÑÑедÑÑÐ²Ð°Ñ Ð±ÐµÐ· водиÑелÑ. ÐÑобе внимание бÑло Ñделено Ñаким вопÑоÑам, как обÑабоÑка видеоконÑенÑа. ÐÑо ÑÑебÑеÑÑÑ Ð´Ð»Ñ Ð¿ÑинÑÑÐ¸Ñ Ð´Ð°Ð½Ð½ÑÑ Ð¾ внеÑней ÑÑеде. ÐÑо даÑÑ Ð²Ð¾Ð·Ð¼Ð¾Ð¶Ð½Ð¾ÑÑÑ Ð² бÑдÑÑем ÑеÑиÑÑ Ð²Ð¾Ð¿ÑоÑÑ Ð¾Ð± ÑпÑавлÑемÑÑ Ð¼Ð°Ð½ÐµÐ²ÑÐ°Ñ Ð°Ð²ÑомобилÑ. Создание полноÑенной ÑиÑÑÐµÐ¼Ñ Ð·ÑÐµÐ½Ð¸Ñ Ð´Ð»Ñ Ð°Ð²ÑÐ¾Ð¼Ð¾Ð±Ð¸Ð»Ñ Ð±ÐµÐ· водиÑелÑ, благодаÑÑ ÐºÐ¾ÑоÑой возможно легкое воÑпÑиÑÑие окÑÑжаÑÑей ÑÑÐµÐ´Ñ Ð¸Ð· видеопоÑока â ÑÐµÐ»Ñ Ð´Ð°Ð½Ð½Ð¾Ð¹ квалиÑикаÑионной ÑабоÑÑ. ÐÑинÑÑие кадÑов из видеопоÑока на Ð²Ñ Ð¾Ð´Ðµ. ÐÑи ÑÑом, локализаÑÐ¸Ñ ÐºÐ°Ð¼ÐµÑÑ Ð´Ð¾Ð»Ð¶Ð½Ð° бÑÑÑ Ð² пеÑедней ÑаÑÑи ÑÑанÑпоÑÑного ÑÑедÑÑва, оÑиенÑаÑÐ¸Ñ â впеÑед, ÑмоÑÑÑ Ð½Ð° меÑÑо пеÑед ÑÑанÑпоÑÑнÑм ÑÑедÑÑвом. Ðа вÑÑ Ð¾Ð´Ðµ ÑиÑÑема бÑÐ´ÐµÑ Ð¸Ð½ÑоÑмиÑоваÑÑ Ð´Ð°Ð½Ð½Ñе ÑÑаÑÑÑ â ÑазмеÑкÑ. Также ÑаÑÑмаÑÑиваеÑÑÑ Ð²Ð°ÑÐ¸Ð°Ð½Ñ ÑÐ¾Ð·Ð´Ð°Ð½Ð¸Ñ ÑÑаекÑоÑии Ð´Ð²Ð¸Ð¶ÐµÐ½Ð¸Ñ Ð¿Ñи иÑÑледовании ÑазмеÑеннÑÑ Ð´Ð°Ð½Ð½ÑÑ ÑоÑÑе. Ð ÑезÑлÑÑаÑе ÑабоÑÑ Ð±Ñли ÑаÑÑмоÑÑÐµÐ½Ñ Ð²Ð¸Ð´Ñ Ð¸ ÑпоÑÐ¾Ð±Ñ Ð¾Ð±ÑабоÑки изобÑажений Ñ ÑелÑÑ Ð¿Ð¾Ð´Ð±Ð¾Ñа ÑÑÑаÑегии компÑÑÑеÑного зÑÐµÐ½Ð¸Ñ Ð¸ маÑинного обÑÑениÑ. ÐÑенка и каÑеÑÑво алгоÑиÑмов пÑовеÑÑлоÑÑ Ð½Ð° видеоÑÐ°Ð¹Ð»Ð°Ñ , на коÑоÑÑе запиÑана ÑÑемка гоноÑной ÑÑаÑÑÑ. ÐапиÑаннÑе даннÑе имеÑÑ Ð²ÑÑокое ÑазÑеÑение, ÑÑо позволÑÐµÑ Ð´ÐµÑалÑно ÑаÑÑмоÑÑеÑÑ Ð²Ñе меÑки и дÑÑгие немаловажнÑе моменÑÑ. ÐеÑки, коÑоÑÑе иÑполÑзÑÑÑÑÑ Ð´Ð»Ñ Ð²ÑÐ´ÐµÐ»ÐµÐ½Ð¸Ñ ÑÑаÑÑÑ, доÑÑаÑоÑно Ñ Ð¾ÑоÑо оÑобÑажаÑÑ Ð´Ð¾ÑÐ¾Ð³Ñ Ð´Ð°Ð¶Ðµ пÑи ÑамÑÑ Ð½ÐµÐ±Ð»Ð°Ð³Ð¾Ð¿ÑиÑÑнÑÑ ÐºÐ»Ð¸Ð¼Ð°ÑиÑеÑÐºÐ¸Ñ ÑÑловиÑÑ . ÐÑ Ð¼Ð¾Ð¶Ð½Ð¾ иÑполÑзоваÑÑ Ð² поÑледÑÑÑÐ¸Ñ Ð¿ÑоÑеÑÑÐ°Ñ ÑиÑÑем зÑÐµÐ½Ð¸Ñ Ð² беÑпилоÑнÑÑ ÑÑедÑÑÐ²Ð°Ñ Ð¿ÐµÑедвижениÑ., Qualification work on the topic: "Algorithm for the application of driving techniques and predicting the trajectory of movement". The scientific project reflects the use of algorithms for a computer vision system and a subsection of artificial intelligence - machine learning. Their application is to establish factors for the use of unmanned vehicles or other vehicles without a driver. Particular attention was paid to issues such as video content processing. This is required for the acceptance of data on the external environment. This will provide an opportunity in the future to solve questions about the controlled maneuvers of the car. The creation of a full-fledged vision system for a car without a driver, thanks to which it is possible to easily perceive the environment from a video stream, is the goal of this qualification work. Receiving frames from a video stream at the input. In this case, the localization of the camera should be in front of the vehicle, orientation - forward, looking at the place in front of the vehicle. At the output, the system will inform the track data - marking. We also consider the option of creating a trajectory when examining the marked highway data. As a result of the work, the types and methods of image processing were considered in order to select a strategy for computer vision and machine learning. The evaluation and quality of the algorithms was checked on video files on which the shooting of the race track was recorded. The recorded data has a high resolution, which allows you to view in detail all the marks and other important points. The marks that are used to highlight the road show the road reasonably well even under the most unfavorable climatic conditions. They can be used in subsequent processes of vision systems in unmanned vehicles.
- Published
- 2021
- Full Text
- View/download PDF
34. OpenTraj: Assessing Prediction Complexity in Human Trajectories Datasets
- Author
-
Javad Amirian, Bingqing Zhang, Francisco Valente Castro, Juan Jose Baldelomar, Jean-Bernard Hayet, Julien Pettré, Sensor-based and interactive robotics (RAINBOW), Inria Rennes – Bretagne Atlantique, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-SIGNAUX ET IMAGES NUMÉRIQUES, ROBOTIQUE (IRISA-D5), Institut de Recherche en Informatique et Systèmes Aléatoires (IRISA), Université de Bretagne Sud (UBS)-Institut National des Sciences Appliquées - Rennes (INSA Rennes), Institut National des Sciences Appliquées (INSA)-Université de Rennes (UNIV-RENNES)-Institut National des Sciences Appliquées (INSA)-Université de Rennes (UNIV-RENNES)-Institut National de Recherche en Informatique et en Automatique (Inria)-École normale supérieure - Rennes (ENS Rennes)-Centre National de la Recherche Scientifique (CNRS)-Université de Rennes 1 (UR1), Université de Rennes (UNIV-RENNES)-CentraleSupélec-IMT Atlantique Bretagne-Pays de la Loire (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT)-Institut Mines-Télécom [Paris] (IMT)-Université de Bretagne Sud (UBS)-Institut National des Sciences Appliquées - Rennes (INSA Rennes), Institut Mines-Télécom [Paris] (IMT)-Institut Mines-Télécom [Paris] (IMT)-Institut de Recherche en Informatique et Systèmes Aléatoires (IRISA), Institut National des Sciences Appliquées (INSA)-Université de Rennes (UNIV-RENNES)-Institut National des Sciences Appliquées (INSA)-Université de Rennes (UNIV-RENNES)-École normale supérieure - Rennes (ENS Rennes)-Centre National de la Recherche Scientifique (CNRS)-Université de Rennes 1 (UR1), Institut Mines-Télécom [Paris] (IMT)-Institut Mines-Télécom [Paris] (IMT), University College of London [London] (UCL), Centro de Investigación en Matemáticas (CIMAT), Consejo Nacional de Ciencia y Tecnología [Mexico] (CONACYT), This research is supported by the CrowdBot H2020 EU Project http://crowdbot.eu/ and by the Intel Probabilistic Computing initiative. The work done by Francisco Valente Castro was sponsored using an MSc Scholarship given by CONACYT with the following scholar registry number 1000188., Université de Rennes (UR)-Institut National des Sciences Appliquées - Rennes (INSA Rennes), Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Université de Bretagne Sud (UBS)-École normale supérieure - Rennes (ENS Rennes)-Institut National de Recherche en Informatique et en Automatique (Inria)-CentraleSupélec-Centre National de la Recherche Scientifique (CNRS)-IMT Atlantique (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT)-Institut Mines-Télécom [Paris] (IMT)-Université de Rennes (UR)-Institut National des Sciences Appliquées - Rennes (INSA Rennes), and Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Université de Bretagne Sud (UBS)-École normale supérieure - Rennes (ENS Rennes)-CentraleSupélec-Centre National de la Recherche Scientifique (CNRS)-IMT Atlantique (IMT Atlantique)
- Subjects
FOS: Computer and information sciences ,0209 industrial biotechnology ,Source code ,Computer science ,media_common.quotation_subject ,Computer Vision and Pattern Recognition (cs.CV) ,Computer Science - Computer Vision and Pattern Recognition ,motion prediction ,Context (language use) ,02 engineering and technology ,Machine learning ,computer.software_genre ,01 natural sciences ,trajectory dataset ,020901 industrial engineering & automation ,Motion prediction ,0103 physical sciences ,[INFO.INFO-RB]Computer Science [cs]/Robotics [cs.RO] ,benchmarking ,Predictability ,trajectory forecasting ,010306 general physics ,media_common ,Series (mathematics) ,business.industry ,Human trajectory prediction ,Benchmarking ,Key (cryptography) ,Trajectory ,dataset assessment ,Artificial intelligence ,business ,computer - Abstract
Human Trajectory Prediction (HTP) has gained much momentum in the last years and many solutions have been proposed to solve it. Proper benchmarking being a key issue for comparing methods, this paper addresses the question of evaluating how complex is a given dataset with respect to the prediction problem. For assessing a dataset complexity, we define a series of indicators around three concepts: Trajectory predictability; Trajectory regularity; Context complexity. We compare the most common datasets used in HTP in the light of these indicators and discuss what this may imply on benchmarking of HTP algorithms. Our source code is released on Github., Comment: ACCV2020
- Published
- 2020
35. AC-VRNN: Attentive Conditional-VRNN for Multi-Future Trajectory Prediction
- Author
-
Simone Calderara, Alessia Bertugli, Lamberto Ballan, Rita Cucchiara, and Pasquale Coscia
- Subjects
FOS: Computer and information sciences ,Computer Science - Machine Learning ,Time series ,Computer science ,Computer Vision and Pattern Recognition (cs.CV) ,Computer Vision ,Computer Science - Computer Vision and Pattern Recognition ,Graph attention networks ,Multi-future prediction ,Trajectory forecasting ,Variational recurrent neural networks ,Machine learning ,computer.software_genre ,Trajectory Prediction ,Machine Learning (cs.LG) ,Machine Learning ,Component (UML) ,Forcing (recursion theory) ,Intersection (set theory) ,business.industry ,Computer Vision, Machine Learning, Trajectory Prediction ,Recurrent neural network ,Signal Processing ,Trajectory ,Graph (abstract data type) ,Computer Vision and Pattern Recognition ,Artificial intelligence ,business ,computer ,Software - Abstract
Anticipating human motion in crowded scenarios is essential for developing intelligent transportation systems, social-aware robots and advanced video surveillance applications. A key component of this task is represented by the inherently multi-modal nature of human paths which makes socially acceptable multiple futures when human interactions are involved. To this end, we propose a generative architecture for multi-future trajectory predictions based on Conditional Variational Recurrent Neural Networks (C-VRNNs). Conditioning mainly relies on prior belief maps, representing most likely moving directions and forcing the model to consider past observed dynamics in generating future positions. Human interactions are modeled with a graph-based attention mechanism enabling an online attentive hidden state refinement of the recurrent estimation. To corroborate our model, we perform extensive experiments on publicly-available datasets (e.g., ETH/UCY, Stanford Drone Dataset, STATS SportVU NBA, Intersection Drone Dataset and TrajNet++) and demonstrate its effectiveness in crowded scenes compared to several state-of-the-art methods., Accepted at Computer Vision and Image Understanding (CVIU)
- Published
- 2020
36. Hybrid machine learning model for action recognition and trajectory forecasting
- Subjects
Trajectory forecasting ,Movement modeling and representation ,�������������������������� ������ ������������������������ �������������� ,�������������������� �������������� ,Action recognition ,���������������� �������������� - Abstract
Collaborative processes where humans and robot agents work together are deployed in professional work-places to enhance effectiveness and productivity. The workers are supported by these robot agents not only to perform better professionally, but also to be able to obtain ergonomically green postures. On the other side, cultural institutions are trying to enhance the cultural experience they offer, engage their audience and also preserve the know-how by adding interactive installations to their exhibitions. The museum visitor has the chance to get into the "real" environment of the craftsman and experience the process of artifact creation. Many cultural institutions provide also with professional training to people that are interested in professions that tend to extinct. For the accomplishment of the goals presented above, gesture recognition is both an essential element and also a challenging task. This task was the main part of this master thesis. For the implementation of the gesture recognition model, a methodology was first developed to visualize the dependencies of the human body parts, which were then mathematically represented and worked as a formula for exploiting previously recorded professional gestures. More specifically, a biomechanical model was created, named the Gesture Operational Model (GOM). This model consists of four assumptions that depict the relationships of the human body parts and describe how movements in general and especially gestures are performed. The State-Space model is used for the mathematical representation of those body part relationships -defined at the GOM-, the coefficients of which, are estimated with the Maximum Likelihood Estimation, via the Kalman filtering method. This process leads to the computation of forecasted values of the gestural time-series. Apart from this, Hidden Markov Models (HMMs) are trained with gestures of an industrial and cultural context. In the recognition phase, a cross-check takes place, when the likelihoods of the gesture recognition engine, don't appear to be confident. In those cases, the gesture recognition results are combined with the forecast performed using the Kalman filtering method, as the use of that method allows to not only recognise classes of gestures but also recognize how good a trajectory has been depicted when compared to the real one. To test the performance of the algorithm, two different datasets were recorded and used, a cultural and an industrial one. In both cases, the industrial worker or the artist is asked to perform their everyday routine. Their gestures were captured, analysed, and features were extracted and used as an input in the gesture recognition engine. The results of gesture recognition with the hybrid method of HMMs combined with the State-Space estimation surpassed the results of the HMMs alone and proved to be very satisfying with a good potential for further research. Those results, along with the forecasting ability of the model, are the elements that make up the scientific contribution of this thesis., ���������������������� ���������������������� ������������ ���������������� ������ ������������������ ������������ �������������������������� ���� ������������ ���������������� ������ ������ ���������������� ������ �������������������������������������� ������ ������ ������������������������������. ���� ���������������������� ���������������������������� ������ ������������ �������� ������������������ ������������ ������ �������� ������ ������ ���������������� �������������� ��������, �������� ������ ������ ���� ������������������ ������ �������������������� �������������� ���� ������ �������������������� ���������������� ���������� ������ ��������������. ��������������������, ���� ���������������������� ���������������� �������������������� ���� �������������������� ������ ���������������������� ���������������� ������ ��������������������, ���� ������������������������ ������ ���������� ������ ������������ ���� ������������������������ ������ ������������������ ������ ������������������������, ������������������������ ������������������������ �������������������������� �������� ���������������� ��������. �� �������������������� ������ ���������������� �������� ������ ���������������� ���� �������� ������ "��������������������" �������������������� ������ �������������� ������ ���� ������������ ���� �������������������� ���������������������� ������������������������. ���������� ���������������������� ���������������� ���������������� ������������ �������������������������� ������������������ ���� ���������� ������ �������������������������� ������ ���������������������� ������ �������������� ���� ������������������������. ������ ������ ���������������� ������ ������������ ������ ���������������������������� ����������������, �� �������������������� ���������������������� ���������� ������ ������������ �������������������������� ������ ���������������� ������ ���� ������������ ���������������� ���������� ������ ������������������������. ������ ������ ���������������� ������ ���������������� ���������������������� ����������������������, ���������������������� ������������ ������ ���������������������� ������ ������ �������������������������� ������ �������������������� ������ ���������������� ������ �������������������� ��������������, ���� ���������� ������ ���������������� ������������������������������ �������������������� ������ �������������������������������� ������ ���� ������������������������ ������ ������������������������ ������ ���������������������������� ���������������������� ������ ���������� ������������������������ ��������������������. ������ ������������������������, �������������������������� ������ ���������������������� ��������������, ���� ���������� �������������������� Gesture Operational Model (GOM). �������� ���� �������������� ���������������������� ������ ���������������� ������������������ ������ ������������������������ ������ �������������� ������ ���������� ������ �������������������� �������������� ������ ���������������������� ������ ���������� ���� ������ ���������� ���������������� ���� ���������������� ������������ ������ ������������ ���� ����������������������. ���� �������������� State-Space ������������������������������ ������ ���� �������������������� ������������������������ ���������� ������ �������������� -�������� ������������������������ ������ ���� GOM-, ���� ���������������������� ������ ������������, �������������������� ���� ���� ������������ ������������������ ���������������� ��������������������������, �������� ������ �������������� Kalman. �������� �� �������������������� ������������ �������� �������������������� ������ �������������������������� ���������� ������ �������������������������� ����������������������. ���������� ������ ��������, ���� ���������� �������������� Markov (HMMs) �������������������������� ���� ���������������������� ������������������������ ������ ������������������������ ������������������. ������ �������� ����������������������, �������������������������������� ��������������, �������� �������� ���� ���������������������� ������ �������������� ���������������������� ����������������������, ������ ���������� ����������������������������, ���� ������������������������ ������������������������ ���� ������ ���������������� ������ �������������������� ������������������������������ ���� ������������ ���������������������������� Kalman. �� ���������� ���������� ������ �������������� ������������������ ������ �������� ������ �������������������� �������������������� ���������������������� �������� ������ ������ �������������������� ������ �������� �������� ������ ������������ �������� ������������������������ ���� ���������������� ���� ������ �������������������� . ������ ���� ���������������� �� �������������� ������ ��������������������, ������������������������ ������ �������������������������������� ������ ���������������������� ������������ ������������������, ������ ���������������������� ������ ������ ����������������������. ������ �������� ������ ����������������������, �� ������������������������ �������������� �� �� ���������������������� ���������������� ���� ������������������ ������ �������������������� ������ �������������������������� ��������������. ���� ���������������������� �������� ������������������������, �������������������� ������ ������������������ ���������������������������� ��������, ���� ���������� �������������������������������� ���� �������������� ������ ������������ ���������������������� ����������������������. ���� ������������������������ ������ ���������������������� ���������������������� ���� ������ ���������������� ������������ ������ HMMs �������� ������������������������ ������ ���������������� State-Space ������������������ ���� ������������������������ ������ HMMs �������� �������� �������������������������������� �������� �������� ������ ������������������������ �������� �������������������������� ���� ������ �������� ���������������� ������ ������������������ ������������. �������� ���� ������������������������, �������� ���� ������ ������������������ ������������������ ������ ����������������, ���������� ���� ���������������� ������ ������������������ ������ ������������������������ �������������� ���������� ������ ������������������������ ����������������.
- Published
- 2020
- Full Text
- View/download PDF
37. End-to-End Pedestrian Trajectory Forecasting with Transformer Network.
- Author
-
Yao, Hai-Yan, Wan, Wang-Gen, and Li, Xiang
- Subjects
- *
FORECASTING , *NATURAL language processing , *COMPUTER vision , *PEDESTRIANS - Abstract
Analysis of pedestrians' motion is important to real-world applications in public scenes. Due to the complex temporal and spatial factors, trajectory prediction is a challenging task. With the development of attention mechanism recently, transformer network has been successfully applied in natural language processing, computer vision, and audio processing. We propose an end-to-end transformer network embedded with random deviation queries for pedestrian trajectory forecasting. The self-correcting scheme can enhance the robustness of the network. Moreover, we present a co-training strategy to improve the training effect. The whole scheme is trained collaboratively by the original loss and classification loss. Therefore, we also achieve more accurate prediction results. Experimental results on several datasets indicate the validity and robustness of the network. We achieve the best performance in individual forecasting and comparable results in social forecasting. Encouragingly, our approach achieves a new state of the art on the Hotel and Zara2 datasets compared with the social-based and individual-based approaches. [ABSTRACT FROM AUTHOR]
- Published
- 2022
- Full Text
- View/download PDF
38. STENet: A hybrid spatio-temporal embedding network for human trajectory forecasting.
- Author
-
Zhang, Bo, Yuan, Chengzhi, Wang, Tao, and Liu, Hongbo
- Subjects
- *
CONVOLUTIONAL neural networks , *TIME-varying networks , *FORECASTING - Abstract
In this paper, we present a hybrid spatio-temporal embedding network (named as STENet) for human trajectory forecasting, which is built upon a GAN-based hierarchical framework. Differently from traditional approaches that only use LSTM for trajectory modeling, we exploit the 1D Convolutional Neural Network (1D-CNN) to embed position features at multiple temporal scales. Moreover, we propose a two-stage graph attention mechanism, which can better describe mutual interactions among pedestrians in the crowd. Additionally, group influences at every time step are taken into account as well. The overall framework is designed using a hierarchical manner, and trained using the Wasserstein distance. We carry out our experiments on the ETH and the UCY datasets. The corresponding results demonstrate the effectiveness of the proposed framework. • We present a hybrid spatio-temporal embedding network for human trajectory forecasting. • We combine 1D-CNN and LSTM modules to embed position features at multiple temporal scales. • We exploit a clustering strategy to provide the spatial layout of pedestrians at the group level. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF
39. Forecasting People Trajectories and Head Poses by Jointly Reasoning on Tracklets and Vislets
- Author
-
Alessio Del Bue, Francesco Setti, Sikandar Amin, Vasileios Belagiannis, Fabio Galasso, Marco Cristani, Theodore Tsesmelis, and Irtiza Hasan
- Subjects
FOS: Computer and information sciences ,LSTM, Trajectory Forecasting, head pose estimation, visual attention, gaze estimation ,Exploit ,Computer Vision and Pattern Recognition (cs.CV) ,LSTM ,Trajectory Forecasting ,RNN ,head pose estimation ,visual attention ,gaze estimation ,computer vision ,machine learning ,Computer Science - Computer Vision and Pattern Recognition ,Time horizon ,02 engineering and technology ,Pedestrian ,Artificial Intelligence ,0202 electrical engineering, electronic engineering, information engineering ,Leverage (statistics) ,Joint problem ,business.industry ,Applied Mathematics ,Covariance ,Backpropagation ,Visualization ,Computational Theory and Mathematics ,020201 artificial intelligence & image processing ,Computer Vision and Pattern Recognition ,Artificial intelligence ,business ,Software - Abstract
In this work, we explore the correlation between people trajectories and their head orientations. We argue that people trajectory and head pose forecasting can be modelled as a joint problem. Recent approaches on trajectory forecasting leverage short-term trajectories (aka tracklets) of pedestrians to predict their future paths. In addition, sociological cues, such as expected destination or pedestrian interaction, are often combined with tracklets. In this paper, we propose MiXing-LSTM (MX-LSTM) to capture the interplay between positions and head orientations (vislets) thanks to a joint unconstrained optimization of full covariance matrices during the LSTM backpropagation. We additionally exploit the head orientations as a proxy for the visual attention, when modeling social interactions. MX-LSTM predicts future pedestrians location and head pose, increasing the standard capabilities of the current approaches on long-term trajectory forecasting. Compared to the state-of-the-art, our approach shows better performances on an extensive set of public benchmarks. MX-LSTM is particularly effective when people move slowly, i.e. the most challenging scenario for all other models. The proposed approach also allows for accurate predictions on a longer time horizon., Comment: Accepted at IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 2019. arXiv admin note: text overlap with arXiv:1805.00652
- Published
- 2019
- Full Text
- View/download PDF
40. Head Pose Estimation and Trajectory Forecasting
- Author
-
Hasan, Irtiza
- Subjects
Detection ,Surveillance ,Settore INF/01 - Informatica ,Head Pose Estimation, Trajectory Forecasting, CNN, LSTM, RNN, Detection, Tracking, Surveillance ,Head Pose Estimation ,Tracking ,Trajectory Forecasting ,LSTM ,RNN ,CNN - Published
- 2019
41. AC-VRNN: Attentive Conditional-VRNN for multi-future trajectory prediction.
- Author
-
Bertugli, Alessia, Calderara, Simone, Coscia, Pasquale, Ballan, Lamberto, and Cucchiara, Rita
- Subjects
INTELLIGENT transportation systems ,RECURRENT neural networks ,ARTIFICIAL neural networks ,VIDEO surveillance ,PROBABILISTIC generative models ,FORECASTING ,SOCIAL interaction - Abstract
Anticipating human motion in crowded scenarios is essential for developing intelligent transportation systems, social-aware robots and advanced video surveillance applications. A key component of this task is represented by the inherently multi-modal nature of human paths which makes socially acceptable multiple futures when human interactions are involved. To this end, we propose a generative architecture for multi-future trajectory predictions based on Conditional Variational Recurrent Neural Networks (C-VRNNs). Conditioning mainly relies on prior belief maps, representing most likely moving directions and forcing the model to consider past observed dynamics in generating future positions. Human interactions are modelled with a graph-based attention mechanism enabling an online attentive hidden state refinement of the recurrent estimation. To corroborate our model, we perform extensive experiments on publicly-available datasets (e.g., ETH/UCY, Stanford Drone Dataset, STATS SportVU NBA, Intersection Drone Dataset and TrajNet++) and demonstrate its effectiveness in crowded scenes compared to several state-of-the-art methods. • Multi-future trajectory predictions in crowded scenarios are considered. • We propose a model based on Conditional Variational Recurrent Neural Networks. • Prior belief maps steer predictions mimicking human behaviours. • An attentive-based graph neural network models interactions among pedestrians. • We outperform state-of-the-art methods on several trajectory prediction benchmarks. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF
42. Gaussian Process Regression-based GPS Variance Estimation and Trajectory Forecasting
- Abstract
Spatio-temporal data is a commonly used source of information. Using machine learning to analyse this kind of data can lead to many interesting and useful insights. In this thesis project, a novel public transportation spatio-temporal dataset is explored and analysed. The dataset contains 282 GB of positional events, spanning two weeks of time, from all public transportation vehicles in Östergötland county, Sweden. From the data exploration, three high-level problems are formulated: bus stop detection, GPS variance estimation, and arrival time prediction, also called trajectory forecasting. The bus stop detection problem is briefly discussed and solutions are proposed. Gaussian process regression is an effective method for solving regression problems. The GPS variance estimation problem is solved via the use of a mixture of Gaussian processes. A mixture of Gaussian processes is also used to predict the arrival time for public transportation buses. The arrival time prediction is from one bus stop to the next, not for the whole trajectory. The result from the arrival time prediction is a distribution of arrival times, which can easily be applied to determine the earliest and latest expected arrival to the next bus stop, alongside the most probable arrival time. The naïve arrival time prediction model implemented has a root mean square error of 5 to 19 seconds. In general, the absolute error of the prediction model decreases over time in each respective segment. The results from the GPS variance estimation problem is a model which can compare the variance for different environments along the route on a given trajectory.
- Published
- 2018
43. Gaussian Process Regression-based GPS Variance Estimation and Trajectory Forecasting
- Abstract
Spatio-temporal data is a commonly used source of information. Using machine learning to analyse this kind of data can lead to many interesting and useful insights. In this thesis project, a novel public transportation spatio-temporal dataset is explored and analysed. The dataset contains 282 GB of positional events, spanning two weeks of time, from all public transportation vehicles in Östergötland county, Sweden. From the data exploration, three high-level problems are formulated: bus stop detection, GPS variance estimation, and arrival time prediction, also called trajectory forecasting. The bus stop detection problem is briefly discussed and solutions are proposed. Gaussian process regression is an effective method for solving regression problems. The GPS variance estimation problem is solved via the use of a mixture of Gaussian processes. A mixture of Gaussian processes is also used to predict the arrival time for public transportation buses. The arrival time prediction is from one bus stop to the next, not for the whole trajectory. The result from the arrival time prediction is a distribution of arrival times, which can easily be applied to determine the earliest and latest expected arrival to the next bus stop, alongside the most probable arrival time. The naïve arrival time prediction model implemented has a root mean square error of 5 to 19 seconds. In general, the absolute error of the prediction model decreases over time in each respective segment. The results from the GPS variance estimation problem is a model which can compare the variance for different environments along the route on a given trajectory.
- Published
- 2018
44. 'Seeing is Believing': Pedestrian Trajectory Forecasting Using Visual Frustum of Attention
- Author
-
Francesco Setti, Fabio Galasso, Theodore Tsesmelis, Irtiza Hasan, Alessio Del Bue, and Marco Cristani
- Subjects
Computer science ,forecasting ,02 engineering and technology ,Head Pose Estimation, trajectory forecasting ,pose estimation ,computer vision ,Oracle ,Task (project management) ,03 medical and health sciences ,0302 clinical medicine ,0202 electrical engineering, electronic engineering, information engineering ,Computer vision ,trajectory forecasting ,Pose ,Frustum ,machine learning ,business.industry ,Head Pose Estimation ,Visualization ,Viewing frustum ,Trajectory ,Task analysis ,020201 artificial intelligence & image processing ,Artificial intelligence ,business ,030217 neurology & neurosurgery - Abstract
In this paper we show the importance of the head pose estimation in the task of trajectory forecasting. This cue, when produced by an oracle and injected in a novel socially-based energy minimization approach, allows to get state-of-the-art performances on four different forecasting benchmarks, without relying on additional information such as expected destination and desired speed, which are supposed to be know beforehand for most of the current forecasting techniques. Our approach uses the head pose estimation for two aims: 1) to define a view frustum of attention, highlighting the people a given subject is more interested about, in order to avoid collisions; 2) to give a shorttime estimation of what would be the desired destination point. Moreover, we show that when the head pose estimation is given by a real detector, though the performance decreases, it still remains at the level of the top score forecasting systems.
- Published
- 2018
45. Regression med Gaussiska Processer för Estimering av GPS Varians och Trajektoriebaserade Tidtabellsprognoser
- Author
-
Kortesalmi, Linus
- Subjects
Variance Estimation ,trajektoria ,Maskininlärning ,GPR ,Computer Sciences ,Trajectory ,Gaussian Process Regression ,Trajectory Forecasting ,Regression ,Machine Learning ,Gaussiska Processer ,Datavetenskap (datalogi) ,GP ,Statistik ,Gaussian Process ,Variansestimering - Abstract
Spatio-temporal data is a commonly used source of information. Using machine learning to analyse this kind of data can lead to many interesting and useful insights. In this thesis project, a novel public transportation spatio-temporal dataset is explored and analysed. The dataset contains 282 GB of positional events, spanning two weeks of time, from all public transportation vehicles in Östergötland county, Sweden. From the data exploration, three high-level problems are formulated: bus stop detection, GPS variance estimation, and arrival time prediction, also called trajectory forecasting. The bus stop detection problem is briefly discussed and solutions are proposed. Gaussian process regression is an effective method for solving regression problems. The GPS variance estimation problem is solved via the use of a mixture of Gaussian processes. A mixture of Gaussian processes is also used to predict the arrival time for public transportation buses. The arrival time prediction is from one bus stop to the next, not for the whole trajectory. The result from the arrival time prediction is a distribution of arrival times, which can easily be applied to determine the earliest and latest expected arrival to the next bus stop, alongside the most probable arrival time. The naïve arrival time prediction model implemented has a root mean square error of 5 to 19 seconds. In general, the absolute error of the prediction model decreases over time in each respective segment. The results from the GPS variance estimation problem is a model which can compare the variance for different environments along the route on a given trajectory.
- Published
- 2018
46. Apropiación y uso tecnologías ADS-B en el CETAD
- Abstract
Air transport has been growing over the last few years, and for this reason, maintaining the safety and improving navigation systems support is a constant challenge in congested airspace. An alternative that has been accepted worldwide because of its advantageous costs and benefits, involves the integration of satellite navigation systems with autonomous broadcasting systems on aircraft, which have allowed to extend situational awareness into areas without coverage of CNS/ATM systems, through an automatic dependent surveillance aircraft system called ADS-B network which allows flight profile sharing with nearby aircraft and ground stations indifferently. This system is being integrated with the monitoring and control systems of many countries, in response to the need of maintaining the air safety in increasingly trafficked airspaces, allowing better use of routes and decreasing operational costs of companies. The Colombian Air Force, through its Technological Development Center for Defense CETAD, is working on understanding and adapting this technology to utilize it in military operations., El transporte aéreo ha venido creciendo en los últimos años, motivo por el cual mantener la seguridad y mejorar el sistema de apoyo a la navegación son retos constantes en el espacio aéreo más congestionado. Una alternativa que ha sido aceptada, a nivel mundial, como ventajosa por sus costos y sus prestaciones, implica la integración de los sistemas satelitales de navegación con sistemas de radiodifusión autónomos en las aeronaves, que han permitido extender la alerta situacional a áreas en las que no se cuenta con cobertura de sistemas CNS/ATM a través de un sistema ADS-B de vigilancia automático dependiente de la aeronave que permite compartir su perfil de vuelo, con aeronaves cercanas y con estaciones de tierra indistintamente. Este sistema se está integrando a los sistemas de vigilancia y control de muchos países, como respuesta a la necesidad de mantener la seguridad aérea en espacios aéreos cada vez más transitados, lo que ha permitido una mejor utilización de las rutas y la disminución de costos operacionales a las empresas. La Fuerza Aérea Colombiana a través del Centro de Desarrollo Tecnológico para la Defensa - CETAD, está trabajando en entender y adaptar esta tecnología para poder explotarla en operaciones militares.
- Published
- 2014
47. Apropiación y uso tecnologías ADS-B en el CETAD
- Abstract
Air transport has been growing over the last few years, and for this reason, maintaining the safety and improving navigation systems support is a constant challenge in congested airspace. An alternative that has been accepted worldwide because of its advantageous costs and benefits, involves the integration of satellite navigation systems with autonomous broadcasting systems on aircraft, which have allowed to extend situational awareness into areas without coverage of CNS/ATM systems, through an automatic dependent surveillance aircraft system called ADS-B network which allows flight profile sharing with nearby aircraft and ground stations indifferently. This system is being integrated with the monitoring and control systems of many countries, in response to the need of maintaining the air safety in increasingly trafficked airspaces, allowing better use of routes and decreasing operational costs of companies. The Colombian Air Force, through its Technological Development Center for Defense CETAD, is working on understanding and adapting this technology to utilize it in military operations., El transporte aéreo ha venido creciendo en los últimos años, motivo por el cual mantener la seguridad y mejorar el sistema de apoyo a la navegación son retos constantes en el espacio aéreo más congestionado. Una alternativa que ha sido aceptada, a nivel mundial, como ventajosa por sus costos y sus prestaciones, implica la integración de los sistemas satelitales de navegación con sistemas de radiodifusión autónomos en las aeronaves, que han permitido extender la alerta situacional a áreas en las que no se cuenta con cobertura de sistemas CNS/ATM a través de un sistema ADS-B de vigilancia automático dependiente de la aeronave que permite compartir su perfil de vuelo, con aeronaves cercanas y con estaciones de tierra indistintamente. Este sistema se está integrando a los sistemas de vigilancia y control de muchos países, como respuesta a la necesidad de mantener la seguridad aérea en espacios aéreos cada vez más transitados, lo que ha permitido una mejor utilización de las rutas y la disminución de costos operacionales a las empresas. La Fuerza Aérea Colombiana a través del Centro de Desarrollo Tecnológico para la Defensa - CETAD, está trabajando en entender y adaptar esta tecnología para poder explotarla en operaciones militares.
- Published
- 2014
48. ADS-B Technologies Use and Appropriation at CETAD
49. ADS-B Technologies Use and Appropriation at CETAD
50. ADS-B Technologies Use and Appropriation at CETAD
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.