Forecasting Time-to-Collision from Monocular Video: Feasibility, Dataset, and Challenges

Manglik, Aashi; Weng, Xinshuo; Ohn-Bar, Eshed

doi:10.48550/arxiv.1903.09102

Cited by 9 publications

(12 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, one of the most important tools in risk assessment and collision avoidance for autonomous agents, e.g., robots and autonomous vehicles, is estimating the TTC [3,5,28,31,35,52]. A popular approach to estimate the TTC is using OF [3,5,35,52].…”

Section: Related Workmentioning

confidence: 99%

Consistent Semantic Attacks on Optical Flow

Koren¹,

Talker²,

Dinerstein³

et al. 2021

Preprint

View full text Add to dashboard Cite

We present a novel approach for semantically targeted adversarial attacks on Optical Flow. In such attacks the goal is to corrupt the flow predictions of a specific object category or instance. Usually, an attacker seeks to hide the adversarial perturbations in the input. However, a quick scan of the output reveals the attack. In contrast, our method helps to hide the attacker's intent in the output as well. We achieve this thanks to a regularization term that encourages off-target consistency. We perform extensive tests on leading optical flow models to demonstrate the benefits of our approach in both white-box and black-box settings. Also, we demonstrate the effectiveness of our attack on subsequent tasks that depend on the optical flow.

show abstract

Section: Related Workmentioning

confidence: 99%

Consistent Semantic Attacks on Optical Flow

Koren¹,

Talker²,

Dinerstein³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Such tasks require fast and precise processing of video data, and egocentric cameras installed on a robot or being weared by a human could provide an advantageous view to capture the most important aspects of the performed activities [26,8]. For human-robot interaction, trajectory forecasting is one of the key research tasks [27].…”

Section: Application Areasmentioning

confidence: 99%

“…Although not widely used in everyday life, other devices besides the aforementioned categories can be useful for applications and research on future prediction in egocentric vision. For example, during the collection of the NearCollision dataset [27] a LIDAR sensor was used for capturing scene depth. The CMU dataset [44] captured human-object interactions by using RF-ID tags.…”

Section: Devicesmentioning

confidence: 99%

See 1 more Smart Citation

Predicting the Future from First Person (Egocentric) Vision: A Survey

Rodin¹,

Furnari²,

Mavroedis³

et al. 2021

Preprint

View full text Add to dashboard Cite

Egocentric videos can bring a lot of information about how humans perceive the world and interact with the environment, which can be beneficial for the analysis of human behaviour. The research in egocentric video analysis is developing rapidly thanks to the increasing availability of wearable devices and the opportunities offered by new large-scale egocentric datasets. As computer vision techniques continue to develop at an increasing pace, the tasks related to the prediction of future are starting to evolve from the need of understanding the present. Predicting future human activities, trajectories and interactions with objects is crucial in applications such as human-robot interaction, assistive wearable technologies for both industrial and daily living scenarios, entertainment and virtual or augmented reality. This survey summarises the evolution of studies in the context of future prediction from egocentric vision making an overview of applications, devices, existing problems, commonly used datasets, models and input modalities. Our analysis highlights that methods for future prediction from egocentric vision can have a significant impact in a range of applications and that further research efforts should be devoted to the standardisation of tasks and the proposal of datasets considering real-world scenarios such as the ones with an industrial vocation.

show abstract

“…A S a fundamental vision task, multiple object tracking (MOT) aims to estimate the locations of several targets [1], [2] and identify which of them belong to the same object [3], [4], [5], [6], [7]. Much attention has been drawn due to its numerous practical applications, such as video analysis [8], autonomous driving [9], robots [10], etc. Although prominent progress has been achieved, existing MOT systems still suffer from poor tracking precision and need improvements.…”

Section: Introductionmentioning

confidence: 99%

RelationTrack: Relation-aware Multiple Object Tracking with Decoupled Representation

Yu¹,

Li²,

Han³

et al. 2021

Preprint

View full text Add to dashboard Cite

Existing online multiple object tracking (MOT) algorithms often consist of two subtasks, detection and reidentification (ReID). In order to enhance the inference speed and reduce the complexity, current methods commonly integrate these double subtasks into a unified framework. Nevertheless, detection and ReID demand diverse features. This issue would result in an optimization contradiction during the training procedure. With the target of alleviating this contradiction, we devise a module named Global Context Disentangling (GCD) that decouples the learned representation into detection-specific and ReID-specific embeddings. As such, this module provides an implicit manner to balance the different requirements of these two subtasks. Moreover, we observe that preceding MOT methods typically leverage local information to associate the detected targets and neglect to consider the global semantic relation. To resolve this restriction, we develop a module, referred to as Guided Transformer Encoder (GTE), by combining the powerful reasoning ability of Transformer encoder and deformable attention. Unlike previous works, GTE avoids analyzing all the pixels and only attends to capture the relation between query nodes and a few self-adaptively selected key samples. Therefore, it is computationally efficient. Extensive experiments have been conducted on the MOT16, MOT17 and MOT20 benchmarks to demonstrate the superiority of the proposed MOT framework, namely RelationTrack. The experimental results indicate that RelationTrack has surpassed preceding methods significantly and established a new state-of-the-art performance, e.g., IDF1 of 70.5% and MOTA of 67.2% on MOT20.

show abstract

Forecasting Time-to-Collision from Monocular Video: Feasibility, Dataset, and Challenges

Cited by 9 publications

References 0 publications

Consistent Semantic Attacks on Optical Flow

Consistent Semantic Attacks on Optical Flow

Predicting the Future from First Person (Egocentric) Vision: A Survey

RelationTrack: Relation-aware Multiple Object Tracking with Decoupled Representation

Contact Info

Product

Resources

About