Summary of Sequential Representation Learning Via Static-dynamic Conditional Disentanglement, by Mathieu Cyrille Simon et al.
Sequential Representation Learning via Static-Dynamic Conditional Disentanglement
by Mathieu Cyrille Simon, Pascal Frossard, Christophe De Vleeschouwer
First submitted to arxiv on: 10 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper proposes a self-supervised model for disentangled representation learning in sequential data, specifically in videos. By accounting for causal relationships between time-independent and time-varying factors, the model improves expressivity through Normalizing Flows. The authors introduce a formal definition of these factors and derive sufficient conditions for identifying ground truth factors. A novel constraint is introduced, which can be incorporated into the framework to promote disentanglement. Experiments show that the proposed approach outperforms state-of-the-art techniques in scenarios where scene dynamics are influenced by content. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper looks at how to learn about videos using AI without needing labeled data. It’s like trying to figure out what’s changing and what’s staying the same in a video. The authors create a new way of doing this that takes into account how things change over time. They show that their method is better than others at understanding videos where the action changes depending on what’s happening. |
Keywords
* Artificial intelligence * Representation learning * Self supervised
