BID: Boundary-Interior Decoding for Unsupervised Temporal Action Localization Pre-Trainin
Not on the map. Why?
Abstract (arXiv)
Skeleton-based motion representations are robust for action localization and understanding for their invariance to perspective, lighting, and occlusion, compared with images. Yet, they are often ambiguous and incomplete when taken out of context, even for human annotators. As infants discern gestures before associating them with words, actions can be conceptualized before being grounded with labels. Therefore, we propose the first unsupervised pre-training framework, Boundary-Interior Decoding (BID), that partitions a skeleton-based motion sequence into discovered semantically meaningful pre-action segments. By fine-tuning our pre-training network with a small number of annotated data, we show results out-performing SOTA methods by a large margin.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 43 of its 57 references that are in the map. First posted on arXiv 12 Mar 2024.
Builds on
43 of 57 references.
References · 43
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
- 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
- 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,750
- 2017Neural Discrete Representation Learningcited by 4,712
- 2017Quo Vadis, Action Recognition? A New Model and the Kinetics Datasetcited by 3,270
- 2020Exploring Simple Siamese Representation Learningcited by 2,866
- 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,857
- 2017Non-local Neural Networkscited by 2,293
- 2020Improved Baselines with Momentum Contrastive Learningcited by 2,287
- 2018SlowFast Networks for Video Recognitioncited by 1,867
- 2018Unsupervised Feature Learning via Non-Parametric Instance-level Discriminationcited by 1,845
31 more
- 2019AMASS: Archive of Motion Capture as Surface Shapescited by 1,236
- 2018Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognitioncited by 1,209
- 2016Temporal Segment Networks: Towards Good Practices for Deep Action Recognitioncited by 902
- 2018TSM: Temporal Shift Module for Efficient Video Understandingcited by 856
- 2021SoundStream: An End-to-End Neural Audio Codeccited by 771
- 2022Autoregressive Image Generation using Residual Quantizationcited by 722
- 2023T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representationscited by 588
- 2023MotionGPT: Human Motion as a Foreign Languagecited by 527
- 2020Spatiotemporal Contrastive Video Representation Learningcited by 372
- 2019BMN: Boundary-Matching Network for Temporal Action Proposal Generationcited by 329
- 2021BABEL: Bodies, Action and Behavior with English Labelscited by 265
- 2019G-TAD: Sub-Graph Localization for Temporal Action Detectioncited by 249
- 2018Videos as Space-Time Region Graphscited by 221
- 2017R-C3D: Region Convolutional 3D Network for Temporal Activity Detectioncited by 161
- 2016Spatio-Temporal LSTM with Trust Gates for 3D Human Action Recognitioncited by 143
- 2020Self-supervised Video Representation Learning by Pace Predictioncited by 139
- 2017A New Representation of Skeleton Sequences for 3D Action Recognitioncited by 126
- 2020Actor-Context-Actor Relation Network for Spatio-Temporal Action Localizationcited by 91
- 2021Contrastive Learning from Extremely Augmented Skeleton Sequences for Self-supervised Action Recognitioncited by 90
- 2023Positive-Augmented Contrastive Learning for Image and Video Captioning Evaluationcited by 78
- 2020TSP: Temporally-Sensitive Pretraining of Video Encoders for Localization Taskscited by 77
- 2021CoLA: Weakly-Supervised Temporal Action Localization with Snippet Contrastive Learningcited by 69
- 2023MoConVQ: Unified Physics-Based Motion Control via Scalable Discrete Representationscited by 60
- 2020Video Representation Learning with Visual Tempo Consistencycited by 57
- 2023Actionlet-Dependent Contrastive Learning for Unsupervised Skeleton-Based Action Recognitioncited by 42
- 2020Boundary-sensitive Pre-training for Temporal Localization in Videoscited by 41
- 2023On the Benefits of 3D Pose and Tracking for Human Action Recognitioncited by 32
- 2022Unsupervised Pre-training for Temporal Action Localization Taskscited by 31
- 2021Foreground-Action Consistency Network for Weakly Supervised Temporal Action Localizationcited by 26
- 2023Ego-Vehicle Action Recognition based on Semi-Supervised Contrastive Learningcited by 8
- 2017Temporal Convolution Based Action Proposal: Submission to ActivityNet 2017cited by 6
Built on by
Citing papers per year: 2025: 2.
2 papers.