LVBench
Emerging8papers using it
2023first seen
LVBench is a benchmark dataset designed to evaluate the performance of models on long video understanding tasks by providing a diverse set of video clips and associated queries.
Papers using LVBench (8)
- Native Active Perception as Reasoning for Omni-Modal UnderstandingSmall Vision-Language Models are Smart Compressors for Long Video UnderstandingQuestion-guided Visual Compression with Memory Feedback for Long-Term Video UnderstandingLensWalk: Agentic Video Understanding by Planning How You See in VideosMSJoE: Jointly Evolving MLLM and Sampler for Efficient Long-Form Video UnderstandingHierarchical Long Video Understanding with Audiovisual Entity Cohesion and Agentic SearchEnhancing Long Video Question Answering With Scene-localized Frame GroupingLvBench: A Benchmark for Long-form Video Understanding with Versatile Multi-modal Question Answering