Instruction-Tuned From survey PDF — verify links
Holmes-VAU
Holmes-VAU — Any-Granularity Understanding
CVPR · 2025Human-reviewed summary status
Survey summary
Holmes-VAU introduces HIVAU-70k, a hierarchical clip/event/video-level instruction benchmark built via a semi-automated annotation engine, and an Anomaly-focused Temporal Sampler that combines an anomaly scorer with density-aware sampling to direct a multimodal LLM toward anomaly-rich segments in long videos.
Taxonomy placement
Instruction-Tuned
Instruction-TunedAny-Granularity Understanding
Method characteristics
Instruction-TunedHIVAU-70KAnomaly-focused samplerHierarchical