Understanding Long Videos with Multimodal Language Models

Benchmark Model Rank Results
zero-shot-video-question-answer-on-egoschema-subsetMVU (13B)#8Accuracy: 60.3Inference Speed (s): 2.42
zero-shot-video-question-answer-on-next-qaMVU (13B)#22Accuracy: 55.2
zero-shot-video-question-answer-on-egoschema-fullsetMVU (13B)#47Accuracy: 37.6