Video-Touch Question Answering on TouchQA (test)
79.3AccuracyCREMA
Evaluation Results
| Method | Links | |
|---|---|---|
| CREMAModality=V, T, N, Tr. Params.=20M, Fine-tuning=true2024.02 | 79.3 | |
| CREMAModality=V, T, Tr. Params.=8M, Fine-tuning=true2024.02 | 79.1 | |
| BLIP-2Modality=V, Tr. Params.=108M, Fine-tuning=true2024.02 | 78.2 | |
| CREMAModality=V, Tr. Params.=4M, Fine-tuning=true2024.02 | 78 | |
| BLIP-2Modality=V, T, Tr. Params.=216M, Fine-tuning=true2024.02 | 77.4 | |
| CMCModality=V, T, Tr. Params.=12M, Fine-tuning=true2024.02 | 44.3 |