Split and Connect: A Universal Tracklet Booster for Multi-Object Tracking

Wang, Gaoang; Wang, Yizhou; Gu, Renshu; Hu, Weijie; Hwang, Jenq‐Neng

doi:10.1109/tmm.2022.3140919

Cited by 32 publications

(16 citation statements)

References 88 publications

(86 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[40], [182], [186] achieve top performance on MOT17, MOT20 and KITTI, respectively. [182] designs a post-processing strategy with tracklet embedding that can be extended to baseline trackers. [186] learns spatial-temporal features for pairwise tracklet embedding.…”

Section: In-depth Analysis On State-of-the-art Methodsmentioning

confidence: 98%

“…[40] proposes a reconstruct-to-embed strategy and employs cross-track relations for tracklet embedding. Since [182], [186] do not refine the public detections, the performance is not competitive with private trackers. To further boost the performance, reliable detections are needed for tracklet embedding methods.…”

Section: In-depth Analysis On State-of-the-art Methodsmentioning

confidence: 99%

“…Then learning the tracklet embedding is similar to the set classification problem. For example, [182] uses multihead attention to aggregate features of individual detections and learns the tracklet embedding with the combination of triplet loss and cross-entropy loss for the association. [106] adopts contrastive loss between detection embedding and tracklet embedding using average features of detections.…”

Section: Self-embeddingmentioning

confidence: 99%

See 2 more Smart Citations

Recent Advances in Embedding Methods for Multi-Object Tracking: A Survey

Wang¹,

Song²,

Hwang³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Multi-object tracking (MOT) aims to associate target objects across video frames in order to obtain entire moving trajectories. With the advancement of deep neural networks and the increasing demand for intelligent video analysis, MOT has gained significantly increased interest in the computer vision community. Embedding methods play an essential role in object location estimation and temporal identity association in MOT. Unlike other computer vision tasks, such as image classification, object detection, re-identification, and segmentation, embedding methods in MOT have large variations, and they have never been systematically analyzed and summarized. In this survey, we first conduct a comprehensive overview with in-depth analysis for embedding methods in MOT from seven different perspectives, including patch-level embedding, single-frame embedding, cross-frame joint embedding, correlation embedding, sequential embedding, tracklet embedding, and cross-track relational embedding. We further summarize the existing widely used MOT datasets and analyze the advantages of existing state-of-the-art methods according to their embedding strategies. Finally, some critical yet under-investigated areas and future research directions are discussed.

show abstract

Section: In-depth Analysis On State-of-the-art Methodsmentioning

confidence: 98%

Section: In-depth Analysis On State-of-the-art Methodsmentioning

confidence: 99%

Section: Self-embeddingmentioning

confidence: 99%

See 1 more Smart Citation

Recent Advances in Embedding Methods for Multi-Object Tracking: A Survey

Wang¹,

Song²,

Hwang³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…These learning strategies are also successfully extended to video representation learning, such as [69,2,60,7]. With much progress made recently, visual representation learning is employed in many real-world applications, such as anomaly detection [77,76], and human-based perception [61,63,62,65]. However, the complex hierarchical relationships among instances are seldom explored in the existing works.…”

Section: Related Workmentioning

confidence: 99%

Hierarchical Semi-Supervised Contrastive Learning for Contamination-Resistant Anomaly Detection

Wang¹,

Zhan²,

Wang³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Anomaly detection aims at identifying deviant samples from the normal data distribution. Contrastive learning has provided a successful way to sample representation that enables effective discrimination on anomalies. However, when contaminated with unlabeled abnormal samples in training set under semi-supervised settings, current contrastive-based methods generally 1) ignore the comprehensive relation between training data, leading to suboptimal performance, and 2) require fine-tuning, resulting in low efficiency. To address the above two issues, in this paper, we propose a novel hierarchical semi-supervised contrastive learning (HSCL) framework, for contamination-resistant anomaly detection. Specifically, HSCL hierarchically regulates three complementary relations: sample-to-sample, sample-to-prototype, and normal-to-abnormal relations, enlarging the discrimination between normal and abnormal samples with a comprehensive exploration of the contaminated data. Besides, HSCL is an end-to-end learning approach that can efficiently learn discriminative representations without fine-tuning. HSCL achieves state-of-the-art performance in multiple scenarios, such as one-class classification and cross-dataset detection. Extensive ablation studies further verify the effectiveness of each considered relation. The code is available at https://github.com/GaoangW/HSCL.

show abstract

“…According to the association manners, the tracking-by-detection methods can be categorized into batch and online methods. Batch-based MOT methods [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 ] exploit the detection results of all frames. They can build longer tracks under occlusions and with incomplete detections since they can achieve (temporal) global associations between long frames.…”

Section: Introductionmentioning

confidence: 99%

Effective Multi-Object Tracking via Global Object Models and Object Constraint Learning

Yoo

Lee

Bae

2022

Sensors

View full text Add to dashboard Cite

Effective multi-object tracking is still challenging due to the trade-off between tracking accuracy and speed. Because the recent multi-object tracking (MOT) methods leverage object appearance and motion models so as to associate detections between consecutive frames, the key for effective multi-object tracking is to reduce the computational complexity of learning both models. To this end, this work proposes global appearance and motion models to discriminate multiple objects instead of learning local object-specific models. In concrete detail, it learns a global appearance model using contrastive learning between object appearances. In addition, we learn a global relation motion model using relative motion learning between objects. Moreover, this paper proposes object constraint learning for improving tracking efficiency. This study considers the discriminability of the models as a constraint, and learns both models when inconsistency with the constraint occurs. Therefore, object constraint learning differs from the conventional online learning for multi-object tracking which updates learnable parameters per frame. This work incorporates global models and object constraint learning into the confidence-based association method, and compare our tracker with the state-of-the-art methods on public available MOT Challenge datasets. As a result, we achieve 64.5% MOTA (multi-object tracking accuracy) and 6.54 Hz tracking speed on the MOT16 test dataset. The comparison results show that our methods can contribute to improve tracking accuracy and tracking speed together.

show abstract

Split and Connect: A Universal Tracklet Booster for Multi-Object Tracking

Cited by 32 publications

References 88 publications

Recent Advances in Embedding Methods for Multi-Object Tracking: A Survey

Recent Advances in Embedding Methods for Multi-Object Tracking: A Survey

Hierarchical Semi-Supervised Contrastive Learning for Contamination-Resistant Anomaly Detection

Effective Multi-Object Tracking via Global Object Models and Object Constraint Learning

Contact Info

Product

Resources

About