Video Question Answering on MSR-VTT
94.4AccuracyInternVideo2_s2 6B
Evaluation Results
| Method | Links | |
|---|---|---|
| InternVideo2_s2 6BZero-shot=true2024.03 | 94.4 | |
| InternVideoZero-shot=true2024.03 | 93.4 | |
| VideoCLIPEvaluation Protocol=Fine-tuned2021.09 | 92.1 | |
| VIOLETZero-shot=true2024.03 | 91.9 | |
| VLMEvaluation protocol=Pairwise Matching2021.05 | 91.64 | |
| ClipBERTEvaluation Protocol=Supervised2021.09 | 88.2 | |
| ActBERTEvaluation Protocol=Supervised2021.09 | 85.7 | |
| ActBERTEvaluation protocol=Pairwise Matching2021.05 | 85.7 | |
| JSFusionEvaluation Protocol=Supervised2021.09 | 83.4 | |
| JSFusionEvaluation protocol=Joint Retrieval2021.05 | 83.4 | |
| MLBEvaluation Protocol=Supervised2021.09 | 76.1 | |
| VideoCLIPEvaluation Protocol=Zero-shot2021.09 | 73.9 | |
| VSE-LSTMEvaluation Protocol=Supervised2021.09 | 67.3 | |
| CT-SANEvaluation Protocol=Supervised2021.09 | 66.4 | |
| EITanqueEvaluation Protocol=Supervised2021.09 | 65.5 | |
| SNUVLEvaluation Protocol=Supervised2021.09 | 65.4 | |
| C+LSTM+SA-FC7Evaluation Protocol=Supervised2021.09 | 60.2 | |
| PaLI-3number of frames=162023.10 | 49.3 | |
| mPLUG-2number of frames=162023.10 | 48 | |
| Flamingo2023.05 | 47.4 | |
| FlamingoVideo-language pretraining=true2023.12 | 47.4 | |
| PaLI-XNumber of frames=162023.05 | 47.1 | |
| PaLI-XNumber of frames=82023.05 | 46.9 | |
| mPLUG2-BVideo-language pretraining=true2023.12 | 46.3 | |
| HiTeAVideo-language pretraining=true2023.12 | 45.4 | |
| RTQVideo-language pretraining=false2023.12 | 42.1 | |
| ALPRO#video_PT=2M, #video_FT=full-shot2022.05 | 42.1 | |
| MV-GPTVideo-language pretraining=true2023.12 | 41.7 | |
| COVGTVideo-language pretraining=true2023.12 | 40 | |
| HQGAVideo-language pretraining=false2023.12 | 38.6 | |
| LSTM-fusionEvaluation Protocol=Supervised2021.09 | 38.3 | |
| IGVVideo-language pretraining=false2023.12 | 38.3 | |
| COVGTVideo-language pretraining=false2023.12 | 38.3 | |
| ATPVideo-language pretraining=false2023.12 | 34.3 | |
| Flamingo-80B#video_PT=27M, #video_FT=8-shot2022.05 | 27.6 | |
| Flamingo-80B#video_PT=27M, #video_FT=4-shot2022.05 | 23.9 | |
| VidIL#video_PT=0, #video_FT=5-shot2022.05 | 21.2 | |
| Flamingo-3B#video_PT=27M, #video_FT=8-shot2022.05 | 19.6 | |
| BLIP_VQA#video_PT=0, #video_FT=0-shot2022.05 | 19.2 | |
| Flamingo-3B#video_PT=27M, #video_FT=4-shot2022.05 | 14.9 | |
| BLIP#video_PT=0, #video_FT=5-shot2022.05 | 0.84 | |
| BLIP#video_PT=0, #video_FT=0-shot2022.05 | 0.55 |