Video Question Answering on MSRVTT-MC (test)
97.8AccuracyVideoDistill
Evaluation Results
| Method | Links | |
|---|---|---|
| VideoDistillPretraining set=WebVid, HD-VILA, EgoCLIP, Pairs=11M2024.04 | 97.8 | |
| HiTeA#PT Data=17M2022.12 | 97.4 | |
| HiTeA#PT Data=5M2022.12 | 97.2 | |
| HD-VILA2021.11 | 97.1 | |
| HD-VILAPretraining set=HD-VILA-100M, Pairs=100M2024.04 | 97.1 | |
| LAVENDER#PT Data=5M2022.12 | 96.6 | |
| Clover#PT Data=5M2022.12 | 95 | |
| SINGULARITY-temporal#PT=17M, #Train Frame=42022.06 | 93.7 | |
| All-in-one#PT Data=283M2022.12 | 92.3 | |
| VideoCLIP#PT=136M, #Train Frame=9602022.06 | 92.1 | |
| SINGULARITY#PT=17M, #Train Frame=12022.06 | 92.1 | |
| VideoClipPre-trained=true2021.11 | 92.1 | |
| VideoCLIPPretraining set=HowTo100M, Pairs=136M2024.04 | 92.1 | |
| SINGULARITY#PT=5M, #Train Frame=12022.06 | 92 | |
| SINGULARITY-temporal#PT=5M, #Train Frame=42022.06 | 92 | |
| Singularity#PT Data=5M2022.12 | 92 | |
| VIOLET#PT Data=183M2022.12 | 91.9 | |
| MERLOT#PT=180M, #Train Frame=52022.06 | 90.9 | |
| MERLOT#PT Data=180M2022.12 | 90.9 | |
| ClipBERT#PT=0.2M, #Train Frame=162022.06 | 88.2 | |
| ClipBERTPre-trained=true2021.11 | 88.2 | |
| ClipBERT#PT Data=0.2M2022.12 | 88.2 | |
| ClipBertPretraining set=COCO, VisGenome, Pairs=2.1M2024.04 | 88.2 | |
| ActBERTPre-trained=true2021.11 | 85.7 | |
| ActBERTPretraining set=HowTo100M2024.04 | 85.7 | |
| JSFusion2021.11 | 83.4 | |
| JSFusion2024.04 | 83.4 | |
| MLB2021.11 | 76.1 | |
| MLB2024.04 | 76.1 | |
| CT-SAN2021.11 | 66.4 | |
| CT-SAN2024.04 | 66.4 |