# Action Segmentation on HOI4D official (test) Source: [Action Segmentation on HOI4D official (test)]() Collection index: [llms.txt]() Data updated: 2026-06-02T11:14:46.788Z Evaluation of Action Segmentation performance on the test split of the HOI4D dataset, reporting Accuracy, Edit distance, and F1 scores at various temporal thresholds (10%, 25%, 50%). 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 segmentation | | Dataset (including any stated split) | HOI4D official (test) | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 26 of 26 | | 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. - [Edit Distance](): Lower is better; unit: SCALAR. - [F1 Score @ 10%](): Higher is better; unit: PERCENT. - [F1 Score @ 25%](): Higher is better; unit: PERCENT. - [F1 Score @ 50%](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | MAMBA4D | 85.5 | Not recorded | 69a125142d603e6de3e3c556 | | X4D-SceneFormer + Mamba4D | 85.5 | Not recorded | 6a1e63dcdbd7150506f854be | | X4D | 84.1 | Not recorded | 69a125142d603e6de3e3c556 | | X4D-SceneFormer | 84.1 | Not recorded | 6a1e63dcdbd7150506f854be | | CrossVideo | 83.7 | Not recorded | 69a125142d603e6de3e3c556 | | PST-Transformer + MotionPDE | 83.21 | Not recorded | 6a1e63dcdbd7150506f854be | | PPTr + MotionPDE | 82.42 | Not recorded | 6a1e63dcdbd7150506f854be | | PPTr+C2P | 81.1 | Not recorded | 68fa6e441e209131cea3edb7 | | PPTr+C2P | 81.1 | Not recorded | 69a125142d603e6de3e3c556 | | PPTr + C2P | 81.1 | Not recorded | 6a1e63dcdbd7150506f854be | | P4Trans+ATA | 79.9 | Not recorded | 69a125142d603e6de3e3c556 | | PPTr + Leaf | 79.4 | Not recorded | 6a1e63dcdbd7150506f854be | | PST-Transformer | 79.22 | Not recorded | 6a1e63dcdbd7150506f854be | | PPTr+VideoMAE | 78.6 | Not recorded | 68fa6e441e209131cea3edb7 | | PPTr+STRL | 78.4 | Not recorded | 68fa6e441e209131cea3edb7 | | PPTr | 77.4 | Not recorded | 68fa6e441e209131cea3edb7 | | PPTr | 77.4 | Not recorded | 69a125142d603e6de3e3c556 | | PPTr | 77.4 | Not recorded | 6a1e63dcdbd7150506f854be | | P4Transformer+C2P | 73.5 | Not recorded | 68fa6e441e209131cea3edb7 | | P4Trans+C2P | 73.5 | Not recorded | 69a125142d603e6de3e3c556 | | P4Transformer | 71.2 | Not recorded | 68fa6e441e209131cea3edb7 | | P4Transformer | 71.2 | Not recorded | 69a125142d603e6de3e3c556 | | P4Transformer | 71.2 | Not recorded | 6a1e63dcdbd7150506f854be | | Asformer | 46.8 | Not recorded | 6a1e63dcdbd7150506f854be | | MS-TCN | 44.2 | Not recorded | 6a1e63dcdbd7150506f854be | | MS-TCN++ | 42.2 | Not recorded | 6a1e63dcdbd7150506f854be | ### MAMBA4D Accuracy: 85.5. Result ID: 69f1c14747f34ce924d83cd1. [Reporting paper]() Recorded evaluation configuration: {"Learning paradigm":"Supervised Learning","Reference":"CVPR2025"}. This record does not assert independent reproduction. ### X4D-SceneFormer + Mamba4D Accuracy: 85.5. Result ID: 6a1ebb21fbc373592c19e35f. [Reporting paper]() Recorded evaluation configuration: {"enhancement":"Mamba4D"}. This record does not assert independent reproduction. ### X4D Accuracy: 84.1. Result ID: 69f1c14747f34ce924d83cd4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning paradigm":"Self-Supervised Learning (End-to-End Fine-Tuning)","Reference":"AAAI2024"}. This record does not assert independent reproduction. ### X4D-SceneFormer Accuracy: 84.1. Result ID: 6a1ebb21fbc373592c19e35e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### CrossVideo Accuracy: 83.7. Result ID: 69f1c14747f34ce924d83cd5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning paradigm":"Self-Supervised Learning (End-to-End Fine-Tuning)","Reference":"ICRA2024"}. This record does not assert independent reproduction. ### PST-Transformer + MotionPDE Accuracy: 83.21. Result ID: 6a1ebb21fbc373592c19e361. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"PST-Transformer"}. This record does not assert independent reproduction. ### PPTr + MotionPDE Accuracy: 82.42. Result ID: 6a1ebb21fbc373592c19e360. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"PPTr"}. This record does not assert independent reproduction. ### PPTr+C2P Accuracy: 81.1. Result ID: 6990224e3a6c314849fe5a3a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Frames":150}. This record does not assert independent reproduction. ### PPTr+C2P Accuracy: 81.1. Result ID: 69f1c14747f34ce924d83cd3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning paradigm":"Self-Supervised Learning (End-to-End Fine-Tuning)","Reference":"CVPR2023"}. This record does not assert independent reproduction. ### PPTr + C2P Accuracy: 81.1. Result ID: 6a1ebb21fbc373592c19e35c. [Reporting paper]() Recorded evaluation configuration: {"enhancement":"C2P"}. This record does not assert independent reproduction. ### P4Trans+ATA Accuracy: 79.9. Result ID: 69f1c14747f34ce924d83cd6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning paradigm":"Adaptation for Pre-Trained Models","Reference":"-"}. This record does not assert independent reproduction. ### PPTr + Leaf Accuracy: 79.4. Result ID: 6a1ebb21fbc373592c19e35b. [Reporting paper]() Recorded evaluation configuration: {"enhancement":"Leaf"}. This record does not assert independent reproduction. ### PST-Transformer Accuracy: 79.22. Result ID: 6a1ebb21fbc373592c19e35d. [Reporting paper]() Recorded evaluation configuration: {"implementation":"our own implementation"}. This record does not assert independent reproduction. ### PPTr+VideoMAE Accuracy: 78.6. Result ID: 6990224e3a6c314849fe5a39. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Frames":150}. This record does not assert independent reproduction. ### PPTr+STRL Accuracy: 78.4. Result ID: 6990224e3a6c314849fe5a38. [Reporting paper]() Recorded evaluation configuration: {"Frames":150}. This record does not assert independent reproduction. ### PPTr Accuracy: 77.4. Result ID: 6990224e3a6c314849fe5a37. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Frames":150}. This record does not assert independent reproduction. ### PPTr Accuracy: 77.4. Result ID: 69f1c14747f34ce924d83cd0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning paradigm":"Supervised Learning","Reference":"ECCV2022"}. This record does not assert independent reproduction. ### PPTr Accuracy: 77.4. Result ID: 6a1ebb21fbc373592c19e35a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### P4Transformer+C2P Accuracy: 73.5. Result ID: 6990224e3a6c314849fe5a36. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Frames":150}. This record does not assert independent reproduction. ### P4Trans+C2P Accuracy: 73.5. Result ID: 69f1c14747f34ce924d83cd2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning paradigm":"Self-Supervised Learning (End-to-End Fine-Tuning)","Reference":"CVPR2023"}. This record does not assert independent reproduction. ### P4Transformer Accuracy: 71.2. Result ID: 6990224e3a6c314849fe5a35. [Reporting paper]() Recorded evaluation configuration: {"Frames":150}. This record does not assert independent reproduction. ### P4Transformer Accuracy: 71.2. Result ID: 69f1c14747f34ce924d83ccf. [Reporting paper]() Recorded evaluation configuration: {"Learning paradigm":"Supervised Learning","Reference":"CVPR2021"}. This record does not assert independent reproduction. ### P4Transformer Accuracy: 71.2. Result ID: 6a1ebb21fbc373592c19e359. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Asformer Accuracy: 46.8. Result ID: 6a1ebb21fbc373592c19e358. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### MS-TCN Accuracy: 44.2. Result ID: 6a1ebb21fbc373592c19e356. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### MS-TCN++ Accuracy: 42.2. Result ID: 6a1ebb21fbc373592c19e357. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.