FogAct: A Real-World Benchmark for Foggy Action Recognition

FogAct addresses a gap in action recognition by focusing on foggy conditions, which reduce visibility and contrast. The dataset is a benchmark for evaluating models in real-world scenarios like surveillance and autonomous driving. It was created by author L1007555652 and the official repository was last updated on May 24, 2026.

Use Cases

• Training models for fog-invariant action recognition based on the described real-world foggy conditions.
• Benchmarking autonomous driving perception systems against reduced visibility and motion blur mentioned in the description.
• Evaluating surveillance system performance in low-contrast, foggy environments as described.
• Developing human-computer interaction systems that are resilient to weather-induced visual degradation.

Strengths

• Explicitly designed for a real-world, understudied condition: foggy action recognition.
• Serves as an official benchmark linked to a research paper, suggesting a defined evaluation protocol.
• Last updated on 2026-05-24, indicating recent maintenance.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset size are unknown, which may limit suitability assessment.
• The description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

L1007555652 on Hugging Face

Collection Method

Method of data gathering is not specified in the provided input.

Freshness

Last updated 2026-05-24 10:28:33.