# Unsupervised Learning of Video Representations via Dense Trajectory Clustering

arXiv: https://arxiv.org/abs/2006.15731 · HTML: https://mlsanity.com/paper/2006.15731/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 Jun 2020.

## Built on by (most cited first, 11 of 11)

- [TCLR: Temporal Contrastive Learning for Video Representation](https://mlsanity.com/paper/2101.07974/) (2021)
- [Self-Supervised Learning for Videos: A Survey](https://mlsanity.com/paper/2207.00419/) (2022)
- [Masked Motion Encoding for Self-Supervised Video Representation Learning](https://mlsanity.com/paper/2210.06096/) (2022)
- [Ranking Info Noise Contrastive Estimation: Boosting Contrastive Learning via Ranked Positives](https://mlsanity.com/paper/2201.11736/) (2022)
- [Long Short View Feature Decomposition via Contrastive Video Representation Learning](https://mlsanity.com/paper/2109.11593/) (2021)
- [A Survey of Generative Techniques for Spatial-Temporal Data Mining](https://mlsanity.com/paper/2405.09592/) (2024)
- [Self-supervised Video Representation Learning with Cross-Stream Prototypical Contrasting](https://mlsanity.com/paper/2106.10137/) (2021)
- [Self-Supervised Learning in Multi-Task Graphs through Iterative Consensus Shift](https://mlsanity.com/paper/2103.14417/) (2021)
- [Large Scale Autonomous Driving Scenarios Clustering with Self-supervised Feature Extraction](https://mlsanity.com/paper/2103.16101/) (2021)
- [Learning a Pedestrian Social Behavior Dictionary](https://mlsanity.com/paper/2212.01426/) (2022)
- [JEDI: Joint Expert Distillation in a Semi-Supervised Multi-Dataset Student-Teacher Scenario for Video Action Recognition](https://mlsanity.com/paper/2308.04934/) (2023)

## Abstract (arXiv)

This paper addresses the task of unsupervised learning of representations for action recognition in videos. Previous works proposed to utilize future prediction, or other domain-specific objectives to train a network, but achieved only limited success. In contrast, in the relevant field of image representation learning, simpler, discrimination-based methods have recently bridged the gap to fully-supervised performance. We first propose to adapt two top performing objectives in this class - instance recognition and local aggregation, to the video domain. In particular, the latter approach iterates between clustering the videos in the feature space of a network and updating it to respect the cluster with a non-parametric classification loss. We observe promising performance, but qualitative analysis shows that the learned representations fail to capture motion patterns, grouping the videos based on appearance. To mitigate this issue, we turn to the heuristic-based IDT descriptors, that were manually designed to encode motion patterns in videos. We form the clusters in the IDT space, using these descriptors as a an unsupervised prior in the iterative local aggregation algorithm. Our experiments demonstrates that this approach outperform prior work on UCF101 and HMDB51 action recognition benchmarks. We also qualitatively analyze the learned representations and show that they successfully capture video dynamics.