# Action Recognition on HMDB51 (3 splits) Source: [Action Recognition on HMDB51 (3 splits)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:29:51.908Z Evaluation of video action recognition performance using the HMDB51 dataset, reporting results derived from its three standard splits. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Action Recognition | | Dataset (including any stated split) | HMDB51 (3 splits) | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 19 of 19 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | ZeroI2V | 83.4 | Not recorded | 68fa74131e209131cea5fa63 | | ST-Adapter | 81.7 | Not recorded | 68fa74131e209131cea5fa63 | | MUFI+fine-tuning | 80.9 | Not recorded | 68fa6b9c1e209131cea2f4bd | | ZeroI2V | 79.9 | Not recorded | 68fa74131e209131cea5fa63 | | SlowOnly-8x8-R101 | 79 | Not recorded | 68fa74131e209131cea5fa63 | | ZeroI2V | 78.5 | Not recorded | 68fa74131e209131cea5fa63 | | ST-Adapter | 77.7 | Not recorded | 68fa74131e209131cea5fa63 | | S3D-G | 75.9 | Not recorded | 68fa6b9c1e209131cea2f4bd | | S3D | 75.9 | Not recorded | 68fa74131e209131cea5fa63 | | LGD-3D | 75.7 | Not recorded | 68fa6b9c1e209131cea2f4bd | | I3D | 74.8 | Not recorded | 68fa74131e209131cea5fa63 | | I3D | 74.5 | Not recorded | 68fa6b9c1e209131cea2f4bd | | R(2+1)D | 74.5 | Not recorded | 68fa6b9c1e209131cea2f4bd | | ZeroI2V | 73.7 | Not recorded | 68fa74131e209131cea5fa63 | | STM | 72.2 | Not recorded | 68fa6b9c1e209131cea2f4bd | | MUFI+linear | 72.2 | Not recorded | 68fa6b9c1e209131cea2f4bd | | ARTNet | 70.9 | Not recorded | 68fa6b9c1e209131cea2f4bd | | VideoPrompt | 66.4 | Not recorded | 68fa74131e209131cea5fa63 | | MUFI | 31 | Not recorded | 68fa6b9c1e209131cea2f4bd | ### ZeroI2V Accuracy: 83.4. Result ID: 69932efa3a6c314849144884. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretrain":"CLIP+K400","Backbone":"ViT-L/14","Evaluation Protocol":"PETL"}. This record does not assert independent reproduction. ### ST-Adapter Accuracy: 81.7. Result ID: 69932efa3a6c314849144880. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretrain":"CLIP+K400","Backbone":"ViT-L/14","Evaluation Protocol":"PETL"}. This record does not assert independent reproduction. ### MUFI+fine-tuning Accuracy: 80.9. Result ID: 698e99d7ffc3901de09f06c0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Pre-train":"Multi-Faceted","Evaluation Protocol":"Linear/Fine-tuning Protocol"}. This record does not assert independent reproduction. ### ZeroI2V Accuracy: 79.9. Result ID: 69932efa3a6c314849144883. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretrain":"CLIP","Backbone":"ViT-L/14","Evaluation Protocol":"PETL"}. This record does not assert independent reproduction. ### SlowOnly-8x8-R101 Accuracy: 79. Result ID: 69932efa3a6c31484914487d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretrain":"Kinetics+OmniSource","Evaluation Protocol":"full fine-tuning"}. This record does not assert independent reproduction. ### ZeroI2V Accuracy: 78.5. Result ID: 69932efa3a6c314849144882. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretrain":"CLIP+K400","Backbone":"ViT-B/16","Evaluation Protocol":"PETL"}. This record does not assert independent reproduction. ### ST-Adapter Accuracy: 77.7. Result ID: 69932efa3a6c31484914487f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretrain":"CLIP+K400","Backbone":"ViT-B/16","Evaluation Protocol":"PETL"}. This record does not assert independent reproduction. ### S3D-G Accuracy: 75.9. Result ID: 698e99d7ffc3901de09f06bc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"BN-Inception","Pre-train":"Kinetics-400","Evaluation Protocol":"Linear/Fine-tuning Protocol"}. This record does not assert independent reproduction. ### S3D Accuracy: 75.9. Result ID: 69932efa3a6c31484914487c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Pretrain":"ImageNet+K400","Evaluation Protocol":"full fine-tuning"}. This record does not assert independent reproduction. ### LGD-3D Accuracy: 75.7. Result ID: 698e99d7ffc3901de09f06be. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-101","Pre-train":"Kinetics-600","Evaluation Protocol":"Linear/Fine-tuning Protocol"}. This record does not assert independent reproduction. ### I3D Accuracy: 74.8. Result ID: 69932efa3a6c31484914487b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Pretrain":"ImageNet+K400","Evaluation Protocol":"full fine-tuning"}. This record does not assert independent reproduction. ### I3D Accuracy: 74.5. Result ID: 698e99d7ffc3901de09f06b9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"BN-Inception","Pre-train":"Kinetics-400","Evaluation Protocol":"Linear/Fine-tuning Protocol"}. This record does not assert independent reproduction. ### R(2+1)D Accuracy: 74.5. Result ID: 698e99d7ffc3901de09f06bb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-34","Pre-train":"Kinetics-400","Evaluation Protocol":"Linear/Fine-tuning Protocol"}. This record does not assert independent reproduction. ### ZeroI2V Accuracy: 73.7. Result ID: 69932efa3a6c314849144881. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Pretrain":"CLIP","Backbone":"ViT-B/16","Evaluation Protocol":"PETL"}. This record does not assert independent reproduction. ### STM Accuracy: 72.2. Result ID: 698e99d7ffc3901de09f06bd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Pre-train":"Kinetics-400","Evaluation Protocol":"Linear/Fine-tuning Protocol"}. This record does not assert independent reproduction. ### MUFI+linear Accuracy: 72.2. Result ID: 698e99d7ffc3901de09f06bf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Pre-train":"Multi-Faceted","Evaluation Protocol":"Linear/Fine-tuning Protocol"}. This record does not assert independent reproduction. ### ARTNet Accuracy: 70.9. Result ID: 698e99d7ffc3901de09f06ba. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"BN-Inception","Pre-train":"Kinetics-400","Evaluation Protocol":"Linear/Fine-tuning Protocol"}. This record does not assert independent reproduction. ### VideoPrompt Accuracy: 66.4. Result ID: 69932efa3a6c31484914487e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Pretrain":"CLIP","Evaluation Protocol":"full fine-tuning"}. This record does not assert independent reproduction. ### MUFI Accuracy: 31. Result ID: 698e99d7ffc3901de09f06c1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Pre-train":"Multi-Faceted","Evaluation Protocol":"Zero-shot Protocol"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.