Search

High-definition (Hi-Def) digital maps are an indispensable automated driving technology that is developing rapidly. There are various commercial or governmental map products in the market. It is notable that the U.S. Department of Transportation (USDOT) map tool allows the user to create MAP and Signal Phase and Timing (SPaT) messages with free access. However, an analysis of the accuracy of this map tool is currently lacking in the literature. This paper provides such an analysis. The analysis manually selects 39 feature points within about 200 meters of the verified point and 55 feature points over longer distances from the verified point. All feature locations are surveyed using GNSS and mapped using the USDOT tool. Different error sources are evaluated to allow assessment of the USDOT map accuracy. In this investigation, The USDOT map tool is demonstrated to achieve 17 centimeters horizontal accuracy, which meets the lane-level map requirement. The maximum horizontal map error is less than 30 centimeters., Comment: 6 pages, 6 figures

This report presents the technical details of our submission to the EPIC-Kitchens-100 Multi-Instance Retrieval Challenge 2022. To participate in the challenge, we designed an ensemble consisting of different models trained with two recently developed relevance-augmented versions of the widely used triplet loss. Our submission, visible on the public leaderboard, obtains an average score of 61.02% nDCG and 49.77% mAP., Comment: Ranked joint 1st place in the Multi-Instance Action Retrieval Challenge organized at EPIC@CVPR2022

In this report, we describe the technical details of our submission for the EPIC-Kitchen-100 action anticipation challenge. Our modelings, the higher-order recurrent space-time transformer and the message-passing neural network with edge learning, are both recurrent-based architectures which observe only 2.5 seconds inference context to form the action anticipation prediction. By averaging the prediction scores from a set of models compiled with our proposed training pipeline, we achieved strong performance on the test set, which is 19.61% overall mean top-5 recall, recorded as second place on the public leaderboard.

Current architectures for multi-modality tasks such as visual question answering suffer from their high complexity. As a result, these architectures are difficult to train and require high computational resources. To address these problems we present a CLIP-based architecture that does not require any fine-tuning of the feature extractors. A simple linear classifier is used on the concatenated features of the image and text encoder. During training an auxiliary loss is added which operates on the answer types. The resulting classification is then used as an attention gate on the answer class selection. On the VizWiz 2022 Visual Question Answering Challenge we achieve 60.15 % accuracy on Task 1: Predict Answer to a Visual Question and AP score of 83.78 % on Task 2: Predict Answerability of a Visual Question., Comment: VizWiz Grand Challenge: Describing Images and Videos Taken by Blind People (CVPR Workshop 2022)

Forecasting future events based on evidence of current conditions is an innate skill of human beings, and key for predicting the outcome of any decision making. In artificial vision for example, we would like to predict the next human action before it happens, without observing the future video frames associated to it. Computer vision models for action anticipation are expected to collect the subtle evidence in the preamble of the target actions. In prior studies recurrence modeling often leads to better performance, the strong temporal inference is assumed to be a key element for reasonable prediction. To this end, we propose a unified recurrence modeling for video action anticipation via message passing framework. The information flow in space-time can be described by the interaction between vertices and edges, and the changes of vertices for each incoming frame reflects the underlying dynamics. Our model leverages self-attention as the building blocks for each of the message passing functions. In addition, we introduce different edge learning strategies that can be end-to-end optimized to gain better flexibility for the connectivity between vertices. Our experimental results demonstrate that our proposed method outperforms previous works on the large-scale EPIC-Kitchen dataset.

Globular clusters contain a unique pulsar population, with many exotic systems that can form only in their dense stellar environments. The leap in sensitivity of the upgraded Giant Metrewave Radio Telescope (uGMRT) in India, especially at low radio frequencies ($<$ 1 GHz) has motivated a new search for radio pulsars in a group of eight Southern globular clusters. We discovered PSR J1835$-$3259B, a 1.83-ms pulsar in NGC 6652; this is in a near-circular wide orbit of 28.7 hr with a low-mass ($ \sim 0.2 \, M_{\rm \odot}$) companion, likely a Helium white dwarf. We derived a 10-year timing solution for this system. We also present measurements of scattering, flux densities and spectral indices for some of the previously known pulsars in these GCs. A significant fraction of the pulsars in these clusters have steep spectral indices. Additionally, we detected eight radio point sources not associated with any known pulsar positions in the radio images. There are four newly identified sources, three in NGC 6652 and one in NGC 6539, and one previously identified source each in NGC 1851, NGC 6440, NGC 6544, and Terzan 5. Surprisingly, our images show that our newly discovered pulsar, PSR J1835$-$3259B, is the brightest pulsar in all GCs we have imaged; like other pulsars with broad profiles (Terzan 5 C and O), its flux density in the radio images is much larger than in its pulsations. This indicates that their pulsed emission is only a fraction of their total emission. The detection of radio sources outside the core radii but well within the tidal radii of these clusters show that future GC surveys should complement the search analysis by using the imaging capability of interferometers, and preferentially synthesize large number of search beams in order to obtain a larger field of view., Comment: 18 pages, 9 figures, 7 tables. Accepted for publication in Astronomy & Astrophysics

Video retrieval using natural language queries has attracted increasing interest due to its relevance in real-world applications, from intelligent access in private media galleries to web-scale video search. Learning the cross-similarity of video and text in a joint embedding space is the dominant approach. To do so, a contrastive loss is usually employed because it organizes the embedding space by putting similar items close and dissimilar items far. This framework leads to competitive recall rates, as they solely focus on the rank of the groundtruth items. Yet, assessing the quality of the ranking list is of utmost importance when considering intelligent retrieval systems, since multiple items may share similar semantics, hence a high relevance. Moreover, the aforementioned framework uses a fixed margin to separate similar and dissimilar items, treating all non-groundtruth items as equally irrelevant. In this paper we propose to use a variable margin: we argue that varying the margin used during training based on how much relevant an item is to a given query, i.e. a relevance-based margin, easily improves the quality of the ranking lists measured through nDCG and mAP. We demonstrate the advantages of our technique using different models on EPIC-Kitchens-100 and YouCook2. We show that even if we carefully tuned the fixed margin, our technique (which does not have the margin as a hyper-parameter) would still achieve better performance. Finally, extensive ablation studies and qualitative analysis support the robustness of our approach. Code will be released at \url{https://github.com/aranciokov/RelevanceMargin-ICMR22}., Comment: Accepted for presentation at International Conference on Multimedia Retrieval (ICMR '22)

Many hand-held or mixed reality devices are used with a single sensor for 3D reconstruction, although they often comprise multiple sensors. Multi-sensor depth fusion is able to substantially improve the robustness and accuracy of 3D reconstruction methods, but existing techniques are not robust enough to handle sensors which operate with diverse value ranges as well as noise and outlier statistics. To this end, we introduce SenFuNet, a depth fusion approach that learns sensor-specific noise and outlier statistics and combines the data streams of depth frames from different sensors in an online fashion. Our method fuses multi-sensor depth streams regardless of time synchronization and calibration and generalizes well with little training data. We conduct experiments with various sensor combinations on the real-world CoRBS and Scene3D datasets, as well as the Replica dataset. Experiments demonstrate that our fusion strategy outperforms traditional and recent online depth fusion approaches. In addition, the combination of multiple sensors yields more robust outlier handling and more precise surface reconstruction than the use of a single sensor. The source code and data are available at https://github.com/tfy14esa/SenFuNet., Comment: Accepted to ECCV 2022. 31 pages, 17 figures, 15 Tables