Long-Term Visual Object Tracking Benchmark

Moudgil, Abhinav; Gandhi, Vineet

doi:10.48550/arxiv.1712.01358

Cited by 12 publications

(44 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The CNN layer has a sigmoid based activation on the outputs. The videos from TLP dataset [16] and Basketball dataset [6] (described in Section 4.1) are used to train the model. We sub-sample from the full trajectory of the video at random frames and create a set of input trajectories each of a fixed length of 512 and use these as single instances of training data for the model.…”

Section: Implementation Detailsmentioning

confidence: 99%

“…Stage Performance dataset: We build a Stage Performance dataset that comprises of two wide-angle recordings of staged performances each of 12 and 10 minutes, respectively. The videos are selected from the Track Long and Prosper (TLP) Dataset [16]. The original recordings were done using a static wide-angle camera covering the entire action in the scene.…”

Section: Datasetsmentioning

confidence: 99%

See 1 more Smart Citation

CineFilter: Unsupervised Filtering for Real Time Autonomous Camera Systems

Achary,

Moorthy,

Javed

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

Learning to mimic the smooth and deliberate camera movement of a human cameraman is an essential requirement for autonomous camera systems. This paper presents a novel formulation for online and real-time estimation of smooth camera trajectories. Many works have focused on global optimization of the trajectory to produce an offline output. Some recent works have tried to extend this to the online setting, but lack either in the quality of the camera trajectories or need large labeled datasets to train their supervised model. We propose two models, one a convex optimization based approach and another a CNN based model, both of which can exploit the temporal trends in the camera behavior. Our model is built in an unsupervised way without any ground truth trajectories and is robust to noisy outliers. We evaluate our models on two different settings namely a basketball dataset and a stage performance dataset and compare against multiple baselines and past approaches. Our models outperform other methods on quantitative and qualitative metrics and produce smooth camera trajectories that are motivated by cinematographic principles. These models can also be easily adopted to run in real-time with a low computational cost, making them fit for a variety of applications.

show abstract

Section: Implementation Detailsmentioning

confidence: 99%

Section: Datasetsmentioning

confidence: 99%

CineFilter: Unsupervised Filtering for Real Time Autonomous Camera Systems

Achary,

Moorthy,

Javed

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…The training datasets used in this paper for SALNet and GPGNet are TLP50 [33], DTB70 [23] and LaSOT [14] 2 dataset which totally contain 120 (50 + 70) and 1400 video sequences, respectively 3 . The LaSOT provide both the BBox and natural language annotations of target object, which is suitable for our natural language guided tracking task.…”

Section: Datasets and Evaluation Criterionmentioning

confidence: 99%

“…https://cis.temple.edu/lasot/3 The baseline method pyMDNet used in this paper is implemented based on PyTorch and pre-trained on two long-term dataset TLP[33] and DTB70 dataset[23] for all our experiments 4. https://github.com/QUVA-Lab/lang-tracker…”

mentioning

confidence: 99%

Describe and Attend to Track: Learning Natural Language guided Structural Representation and Visual Attention for Object Tracking

Wang,

Li,

Yang

et al. 2018

Preprint

View full text Add to dashboard Cite

The tracking-by-detection framework requires a set of positive and negative training samples to learn robust tracking models for precise localization of target objects. However, existing tracking models mostly treat different samples independently while ignores the relationship information among them. In this paper, we propose a novel structureaware deep neural network to overcome such limitations. In particular, we construct a graph to represent the pairwise relationships among training samples, and additionally take the natural language as the supervised information to learn both feature representations and classifiers robustly. To refine the states of the target and re-track the target when it is back to view from heavy occlusion and out of view, we elaborately design a novel subnetwork to learn the target-driven visual attentions from the guidance of both visual and natural language cues. Extensive experiments on five tracking benchmark datasets validated the effectiveness of our proposed method.

show abstract

“…For instance, the most commonly used OTB dataset has an average length of about 20 seconds [35] per clip. Work by Moudgil and Gandhi [23] observed a sharp performance drop when the trackers were evaluated on long sequences. Following works [10,33,21] also make similar observations and suggest that we need alternate ways to evaluate and analyze long term tracking performance.…”

Section: Introductionmentioning

confidence: 99%

Exploring 3 R’s of Long-term Tracking: Re-detection, Recovery and Reliability

Karthik

Moudgil

Gandhi

2020

2020 IEEE Winter Conference on Applications of Computer Vision (WACV)

Self Cite

View full text Add to dashboard Cite

Recent works have proposed several long term tracking benchmarks and highlight the importance of moving towards long-duration tracking to bridge the gap with application requirements. The current evaluation methodologies, however, do not focus on several aspects that are crucial in a long term perspective like Re-detection, Recovery, and Reliability. In this paper, we propose novel evaluation strategies for a more in-depth analysis of trackers from a long-term perspective. More specifically, (a) we test redetection capability of the trackers in the wild by simulating virtual cuts, (b) we investigate the role of chance in the recovery of tracker after failure and (c) we propose a novel metric allowing visual inference on the ability of a tracker to track contiguously (without any failure) at a given accuracy. We present several original insights derived from an extensive set of quantitative and qualitative experiments.

show abstract

Long-Term Visual Object Tracking Benchmark

Cited by 12 publications

References 0 publications

CineFilter: Unsupervised Filtering for Real Time Autonomous Camera Systems

CineFilter: Unsupervised Filtering for Real Time Autonomous Camera Systems

Describe and Attend to Track: Learning Natural Language guided Structural Representation and Visual Attention for Object Tracking

Exploring 3 R’s of Long-term Tracking: Re-detection, Recovery and Reliability

Contact Info

Product

Resources

About