Action Recognition on HMDB51 VISPR
60.6Top-1 AccuracyRaw data
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Raw dataEvaluation Model=GPT-4V2026.03 | 60.6 | 70.1 | |
| OursEvaluation Model=GPT-4V2026.03 | 54.3 | 57.3 | |
| DownsampleEvaluation Model=GPT-4V2026.03 | 51.7 | 61.2 | |
| Obf-WeakBlurEvaluation Model=GPT-4V2026.03 | 50.4 | 69.4 | |
| Raw dataEvaluation Model=Video-LLaVA2026.03 | 50.3 | — | |
| Surrogate model of STPrivacyEvaluation Model=GPT-4V2026.03 | 48.8 | 61.7 | |
| Surrogate model of MPPAREvaluation Model=GPT-4V2026.03 | 48.6 | 62.2 | |
| SelectivePrivacyEvaluation Model=GPT-4V2026.03 | 48.5 | 62.8 | |
| Surrogate model of VITAEvaluation Model=GPT-4V2026.03 | 48.3 | 63.2 | |
| Obf-StrongBlurEvaluation Model=GPT-4V2026.03 | 48.1 | 64.4 | |
| Surrogate model of SPActEvaluation Model=GPT-4V2026.03 | 47.8 | 62.1 | |
| Obf-BlackeningEvaluation Model=GPT-4V2026.03 | 47.6 | 63.8 | |
| OursEvaluation Model=Video-LLaVA2026.03 | 47.6 | — | |
| Surrogate model of BDQEvaluation Model=GPT-4V2026.03 | 47.5 | 62.4 | |
| DownsampleEvaluation Model=Video-LLaVA2026.03 | 42.1 | — | |
| SelectivePrivacyEvaluation Model=Video-LLaVA2026.03 | 35.2 | — | |
| Obf-WeakBlurEvaluation Model=Video-LLaVA2026.03 | 33.2 | — | |
| Surrogate model of BDQEvaluation Model=Video-LLaVA2026.03 | 33.1 | — | |
| Surrogate model of MPPAREvaluation Model=Video-LLaVA2026.03 | 32.6 | — | |
| Surrogate model of SPActEvaluation Model=Video-LLaVA2026.03 | 32.4 | — | |
| Surrogate model of STPrivacyEvaluation Model=Video-LLaVA2026.03 | 32.3 | — | |
| Surrogate model of VITAEvaluation Model=Video-LLaVA2026.03 | 31.6 | — | |
| Obf-StrongBlurEvaluation Model=Video-LLaVA2026.03 | 27.2 | — | |
| Obf-BlackeningEvaluation Model=Video-LLaVA2026.03 | 26.6 | — |