SRTube: Video-Language Pre-Training with Action-Centric Video Tube Features and Semantic Role Labeling
Open paper
Benchmark
Model
Rank
Results
visual-question-answering-on-msrvtt-qa
SRTube (extra data)
–
Accuracy: 0.441
visual-question-answering-on-msvd-qa
SRTube (extra data)
–
Accuracy: 0.460
Rank counts only results with a code link.