Reproduce “Spatial-Temporal Perception with Causal Inference for Naturalistic Driving Action Recognition” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Qing Chang, Wei Dai, Zhihao Shuai, Limin Yu, Yutao Yue
Researchers from multiple Chinese institutions introduce STP (Spatial-Temporal Perception), a novel driver action recognition framework that achieves state-of-the-art performance on benchmark datasets using only RGB input through innovative causal inference and spatial-temporal feature fusion, while maintaining real-time capabilities with just 23.7M parameters and 14.2ms latency.
Key insights
- Causal inference significantly improves action recognition by modeling temporal dependencies
- Single RGB camera input can achieve competitive performance when spatial-temporal features are properly fused
- Multi-scale encoding helps capture both global context and local details effectively
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.