VideoDetective: Clue Hunting via both Extrinsic Query and Intrinsic Relevance for Long Video Understanding

Benchmark Model Rank Results
video-question-answering-on-lvbenchVideoDetective (SeedVL-1.5, 20B)#31Accuracy: 51.3
video-question-answering-on-mlvuVideoDetective (SeedVL-1.5, 20B)#51Accuracy: 63.8