SVLTA: Benchmarking Vision-Language Temporal Alignment via Synthetic Video Situation.

Hao Du, Bo Wu, Yan Lu, Zhendong Mao

Browse the full CVPR paper archive.