Summary of Event Stream Based Human Action Recognition: a High-definition Benchmark Dataset and Algorithms, by Xiao Wang et al.
Event Stream based Human Action Recognition: A High-Definition Benchmark Dataset and Algorithms
by Xiao Wang, Shiao Wang, Pengpeng Shao, Bo Jiang, Lin Zhu, Yonghong Tian
First submitted to arxiv on: 19 Aug 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)
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 This paper proposes a large-scale, high-definition human action recognition dataset based on event cameras, called CeleX-HAR. The dataset consists of 150 action categories and 124,625 video sequences, with various factors such as multi-view, illumination, action speed, and occlusion considered. To facilitate comparison, the authors report over 20 mainstream HAR models for future works. Additionally, they introduce a novel Mamba vision backbone network for event stream-based HAR, called EVMamba, which achieves favorable results across multiple datasets by encoding and mining spatio-temporal information from event streams. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper creates a new dataset to help computers recognize human actions better. The dataset uses special cameras that are inspired by how our brains work. These cameras have advantages over regular cameras, but the existing datasets for training AI models are low resolution and limited. This new dataset has high definition and includes many different actions, making it more challenging and useful for developing better AI models. The authors also introduce a new model called EVMamba that can recognize human actions well using this special type of camera. |
