Search

Due to its widespread applications, human action recognition is one of the most widely studied research problems in Computer Vision. Recent studies have shown that addressing it using multimodal data leads to superior performance as compared to relying on a single data modality. During the adoption of deep learning for visual modelling in the last decade, action recognition approaches have mainly relied on Convolutional Neural Networks (CNNs). However, the recent rise of Transformers in visual modelling is now also causing a paradigm shift for the action recognition task. This survey captures this transition while focusing on Multimodal Human Action Recognition (MHAR). Unique to the induction of multimodal computational models is the process of "fusing" the features of the individual data modalities. Hence, we specifically focus on the fusion design aspects of the MHAR approaches. We analyze the classic and emerging techniques in this regard, while also highlighting the popular trends in the adaption of CNN and Transformer building blocks for the overall problem. In particular, we emphasize on recent design choices that have led to more efficient MHAR models. Unlike existing reviews, which discuss Human Action Recognition from a broad perspective, this survey is specifically aimed at pushing the boundaries of MHAR research by identifying promising architectural and fusion design choices to train practicable models. We also provide an outlook of the multimodal datasets from their scale and evaluation viewpoint. Finally, building on the reviewed literature, we discuss the challenges and future avenues for MHAR., Comment: 23 pages, 5 figures and 3 Tables. To appear in ACM Trans. Multimedia Comput. Commun. Appl.(TOMM) 2024

Offline scheduling in Time Sensitive Networking (TSN) utilizing the Time Aware Shaper (TAS) facilitates optimal deterministic latency and jitter-bounds calculation for Time- Triggered (TT) flows. However, the dynamic nature of traffic in industrial settings necessitates a strategy for adaptively scheduling flows without interrupting existing schedules. Our research identifies critical gaps in current dynamic scheduling methods for TSN and introduces the novel GCN-TD3 approach. This novel approach utilizes a Graph Convolutional Network (GCN) for representing the various relations within different components of TSN and employs the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm to dynamically schedule any incoming flow. Additionally, an Integer Linear Programming (ILP) based offline scheduler is used both to initiate the simulation and serve as a fallback mechanism. This mechanism is triggered to recalculate the entire schedule when the predefined threshold of Gate Control List(GCL) length exceeds. Comparative analyses demonstrate that GCN-TD3 outperforms existing methods like Deep Double Q-Network (DDQN) and Deep Deterministic Policy Gradient (DDPG), exhibiting convergence within 4000 epochs with a 90\% dynamic TT flow admission rate while maintaining deadlines and reducing jitter to as low as 2us. Finally, two modules were developed for the OMNeT++ simulator, facilitating dynamic simulation to evaluate the methodology., Comment: Accepted article IFIP/IEEE Networking 2024 (Tensor)

Traffic state forecasting is crucial for traffic management and control strategies, as well as user- and system-level decision making in the transportation network. While traffic forecasting has been approached with a variety of techniques over the last couple of decades, most approaches simply rely on endogenous traffic variables for state prediction, despite the evidence that exogenous factors can significantly impact traffic conditions. This paper proposes a multi-dimensional spatio-temporal graph attention-based traffic prediction approach (M-STGAT), which predicts traffic based on past observations of speed, along with lane closure events, temperature, and visibility across the transportation network. The approach is based on a graph attention network architecture, which also learns based on the structure of the transportation network on which these variables are observed. Numerical experiments are performed using traffic speed and lane closure data from the California Department of Transportation (Caltrans) Performance Measurement System (PeMS). The corresponding weather data were downloaded from the National Oceanic and Atmospheric Administration (NOOA) Automated Surface Observing Systems (ASOS). For comparison, the numerical experiments implement three alternative models which do not allow for the multi-dimensional input. The M-STGAT is shown to outperform the three alternative models, when performing tests using our primary data set for prediction with a 30-, 45-, and 60-minute prediction horizon, in terms of three error measures: Mean Absolute Error (MAE), Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). However, the model's transferability can vary for different transfer data sets and this aspect may require further investigation., Comment: Transportation Research Board Annual Meeting 2023

In line with the human capacity to perceive the world by simultaneously processing and integrating high-dimensional inputs from multiple modalities like vision and audio, we propose a novel model, MAiVAR-T (Multimodal Audio-Image to Video Action Recognition Transformer). This model employs an intuitive approach for the combination of audio-image and video modalities, with a primary aim to escalate the effectiveness of multimodal human action recognition (MHAR). At the core of MAiVAR-T lies the significance of distilling substantial representations from the audio modality and transmuting these into the image domain. Subsequently, this audio-image depiction is fused with the video modality to formulate a unified representation. This concerted approach strives to exploit the contextual richness inherent in both audio and video modalities, thereby promoting action recognition. In contrast to existing state-of-the-art strategies that focus solely on audio or video modalities, MAiVAR-T demonstrates superior performance. Our extensive empirical evaluations conducted on a benchmark action recognition dataset corroborate the model's remarkable performance. This underscores the potential enhancements derived from integrating audio and video modalities for action recognition purposes., Comment: 6 pages, 7 figures, 4 tables, Peer reviewed, Accepted @ The 11th European Workshop on Visual Information Processing (EUVIP) will be held on 11th-14th September 2023, in Gj{\o}vik, Norway. arXiv admin note: text overlap with arXiv:2103.15691 by other authors

Annotating medical images for disease detection is often tedious and expensive. Moreover, the available training samples for a given task are generally scarce and imbalanced. These conditions are not conducive for learning effective deep neural models. Hence, it is common to 'transfer' neural networks trained on natural images to the medical image domain. However, this paradigm lacks in performance due to the large domain gap between the natural and medical image data. To address that, we propose a novel concept of Pre-text Representation Transfer (PRT). In contrast to the conventional transfer learning, which fine-tunes a source model after replacing its classification layers, PRT retains the original classification layers and updates the representation layers through an unsupervised pre-text task. The task is performed with (original, not synthetic) medical images, without utilizing any annotations. This enables representation transfer with a large amount of training data. This high-fidelity representation transfer allows us to use the resulting model as a more effective feature extractor. Moreover, we can also subsequently perform the traditional transfer learning with this model. We devise a collaborative representation based classification layer for the case when we leverage the model as a feature extractor. We fuse the output of this layer with the predictions of a model induced with the traditional transfer learning performed over our pre-text transferred model. The utility of our technique for limited and imbalanced data classification problem is demonstrated with an extensive five-fold evaluation for three large-scale models, tested for five different class-imbalance ratios for CT based COVID-19 detection. Our results show a consistent gain over the conventional transfer learning with the proposed method., Comment: Best paper at IVCNZ

Currently, action recognition is predominately performed on video data as processed by CNNs. We investigate if the representation process of CNNs can also be leveraged for multimodal action recognition by incorporating image-based audio representations of actions in a task. To this end, we propose Multimodal Audio-Image and Video Action Recognizer (MAiVAR), a CNN-based audio-image to video fusion model that accounts for video and audio modalities to achieve superior action recognition performance. MAiVAR extracts meaningful image representations of audio and fuses it with video representation to achieve better performance as compared to both modalities individually on a large-scale action recognition dataset., Comment: Peer reviewed & accepted at IEEE VCIP 2022 (http://www.vcip2022.org/)

In this paper, we consider a class of structured nonconvex nonsmooth optimization problems, in which the objective function is formed by the sum of a possibly nonsmooth nonconvex function and a differentiable function whose gradient is Lipschitz continuous, subtracted by a weakly convex function. This type of structured problems has many practical applications in machine learning and statistics such as compressed sensing, signal recovery, sparse dictionary learning, clustering, matrix factorization, and others. We develop a flexible extrapolated proximal subgradient algorithm for solving these problems with guaranteed subsequential convergence to a stationary point. The global convergence of the whole sequence generated by our algorithm is also established under the Kurdyka-Lojasiewicz property. To illustrate the promising numerical performance of the proposed algorithm, we conduct numerical experiments on two important nonconvex models. This includes a least squares problem with a nonconvex regularization and an optimal power flow problem with distributed energy resources.

With current trends in sensors (cheaper, more volume of data) and applications (increasing affordability for new tasks, new ideas in what 3D data could be useful for); there is corresponding increasing interest in the ability to automatically, reliably, and cheaply, register together individual point clouds. The volume of data to handle, and still elusive need to have the registration occur fully reliably and fully automatically, mean there is a need to innovate further. One largely untapped area of innovation is that of exploiting the {\em semantic information} of the points in question. Points on a tree should match points on a tree, for example, and not points on car. Moreover, such a natural restriction is clearly human-like - a human would generally quickly eliminate candidate regions for matching based on semantics. Employing semantic information is not only efficient but natural. It is also timely - due to the recent advances in semantic classification capabilities. This paper advances this theme by demonstrating that state of the art registration techniques, in particular ones that rely on "preservation of length under rigid motion" as an underlying matching consistency constraint, can be augmented with semantic information. Semantic identity is of course also preserved under rigid-motion, but also under wider motions present in a scene. We demonstrate that not only the potential obstacle of cost of semantic segmentation, and the potential obstacle of the unreliability of semantic segmentation; are both no impediment to achieving both speed and accuracy in fully automatic registration of large scale point clouds.

Self-excited systems arise in numerous applications, such as biochemical systems, fluid-structure interaction, and combustion. This paper analyzes a discrete-time Lur'e system with a piecewise-linear saturation feedback nonlinearity. The main result provides sufficient conditions under which the Lur'e system is self-excited in the sense that its response is bounded and nonconvergent., Comment: Submitted to Systems & Control Letters, 12 pages, 7 figures

Deep learning is gaining instant popularity in computer aided diagnosis of COVID-19. Due to the high sensitivity of Computed Tomography (CT) to this disease, CT-based COVID-19 detection with visual models is currently at the forefront of medical imaging research. Outcomes published in this direction are frequently claiming highly accurate detection under deep transfer learning. This is leading medical technologists to believe that deep transfer learning is the mainstream solution for the problem. However, our critical analysis of the literature reveals an alarming performance disparity between different published results. Hence, we conduct a systematic thorough investigation to analyze the effectiveness of deep transfer learning for COVID-19 detection with CT images. Exploring 14 state-of-the-art visual models with over 200 model training sessions, we conclusively establish that the published literature is frequently overestimating transfer learning performance for the problem, even in the prestigious scientific sources. The roots of overestimation trace back to inappropriate data curation. We also provide case studies that consider more realistic scenarios, and establish transparent baselines for the problem. We hope that our reproducible investigation will help in curbing hype-driven claims for the critical problem of COVID-19 diagnosis, and pave the way for a more transparent performance evaluation of techniques for CT-based COVID-19 detection., Comment: 9 pages

Within the context of recursive least squares (RLS) parameter estimation, the goal of the present paper is to study the effect of regularization-induced bias on the transient and asymptotic accuracy of the parameter estimates. We consider this question in three stages. First, we consider regression with random data, in which case persistency is guaranteed. Next, we apply RLS to finite-impulse-response (FIR) system identification and, finally, to infinite-impulse-response (IIR) system identification. For each case, we relate the condition number of the regressor matrix to the transient response and rate of convergence of the parameter estimates.