New benchmark tests AI's understanding of long video quality

Original: LongVQUBench: Benchmarking Long-Term Video Quality Understanding of Vision-Language Models

Source: arxiv.org ↗