FAVOR-Bench is a benchmark for fine-grained video motion understanding accepted by NeurIPS 2025. It spans both ego-centric and third-person perspectives and includes evaluation for close-ended QA and open-ended descriptive tasks. The dataset was released by the FAVOR-Bench organization in March 2025.
Use Cases
- Benchmarking video motion understanding models based on fine-grained motion analysis tasks.
- Training multimodal AI systems on tasks requiring perspective recognition (ego-centric vs. third-person).
- Evaluating model performance on both close-ended question answering and open-ended descriptive generation tasks.
Strengths
- Benchmark was accepted by NeurIPS 2025 Datasets and Benchmarks Track.
- Evaluation includes both close-ended QA and open-ended descriptive tasks.
- Dataset spans both ego-centric and third-person video perspectives.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- Freshness should be verified as the last metadata update was 2026-05-11.
Provenance
- Source
- FAVOR-Bench
- Freshness
- Last updated 2026-05-11 08:12:50