# What have we learned from deep representations for action recognition?

arXiv: https://arxiv.org/abs/1801.01415 · HTML: https://mlsanity.com/paper/1801.01415/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 4 Jan 2018.

## Built on by (most cited first, 3 of 3)

- [Self-Supervised Learning for Videos: A Survey](https://mlsanity.com/paper/2207.00419/) (2022)
- [Dogfight: Detecting Drones from Drones Videos](https://mlsanity.com/paper/2103.17242/) (2021)
- [Deep Unsupervised Key Frame Extraction for Efficient Video Classification](https://mlsanity.com/paper/2211.06742/) (2022)

## Abstract (arXiv)

As the success of deep models has led to their deployment in all areas of computer vision, it is increasingly important to understand how these representations work and what they are capturing. In this paper, we shed light on deep spatiotemporal representations by visualizing what two-stream models have learned in order to recognize actions in video. We show that local detectors for appearance and motion objects arise to form distributed representations for recognizing human actions. Key observations include the following. First, cross-stream fusion enables the learning of true spatiotemporal features rather than simply separate appearance and motion features. Second, the networks can learn local representations that are highly class specific, but also generic representations that can serve a range of classes. Third, throughout the hierarchy of the network, features become more abstract and show increasing invariance to aspects of the data that are unimportant to desired distinctions (e.g. motion patterns across various speeds). Fourth, visualizations can be used not only to shed light on learned representations, but also to reveal idiosyncracies of training data and to explain failure cases of the system.