# Few-shot Action Recognition on NTU 120 Source: [Few-shot Action Recognition on NTU 120]() Collection index: [llms.txt]() Data updated: 2026-05-26T10:39:02.180Z Evaluation of few-shot action recognition performance on the NTU 120 dataset, reporting accuracy across varying numbers of classes (e.g., 20, 40, 60, 80, 100 classes). 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 | Few-shot Action Recognition | | Dataset (including any stated split) | NTU 120 | | Metric | Accuracy (20 classes) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 11 of 11 | | 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 (20 classes)](): Higher is better; unit: PERCENT. - [Accuracy (40 classes)](): Higher is better; unit: PERCENT. - [Accuracy (60 classes)](): Higher is better; unit: PERCENT. - [Accuracy (80 classes)](): Higher is better; unit: PERCENT. - [Accuracy (100 classes)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy (20 classes) | Result date | Report paper | | --- | --- | --- | --- | | uDTW | 32.2 | Not recorded | 6a15295bdbd7150506f7e233 | | TAP | 31.2 | Not recorded | 6a15295bdbd7150506f7e233 | | sDTW div. | 30.8 | Not recorded | 6a15295bdbd7150506f7e233 | | sDTW | 30.3 | Not recorded | 6a15295bdbd7150506f7e233 | | uDTW | 22.7 | Not recorded | 6a15295bdbd7150506f7e233 | | ProtoNet | 21.7 | Not recorded | 6a15295bdbd7150506f7e233 | | sDTW div. | 20.8 | Not recorded | 6a15295bdbd7150506f7e233 | | MatchNets | 20.5 | Not recorded | 6a15295bdbd7150506f7e233 | | sDTW | 20.1 | Not recorded | 6a15295bdbd7150506f7e233 | | Euclidean | 18.7 | Not recorded | 6a15295bdbd7150506f7e233 | | Euclidean | 13.5 | Not recorded | 6a15295bdbd7150506f7e233 | ### uDTW Accuracy (20 classes): 32.2. Result ID: 6a1576dffbc373592c0d8c7d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning Setting":"Supervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. ### TAP Accuracy (20 classes): 31.2. Result ID: 6a1576dffbc373592c0d8c79. [Reporting paper]() Recorded evaluation configuration: {"Learning Setting":"Supervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. ### sDTW div. Accuracy (20 classes): 30.8. Result ID: 6a1576dffbc373592c0d8c7c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning Setting":"Supervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. ### sDTW Accuracy (20 classes): 30.3. Result ID: 6a1576dffbc373592c0d8c7b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning Setting":"Supervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. ### uDTW Accuracy (20 classes): 22.7. Result ID: 6a1576dffbc373592c0d8c81. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning Setting":"Unsupervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. ### ProtoNet Accuracy (20 classes): 21.7. Result ID: 6a1576dffbc373592c0d8c78. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning Setting":"Supervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. ### sDTW div. Accuracy (20 classes): 20.8. Result ID: 6a1576dffbc373592c0d8c80. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning Setting":"Unsupervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. ### MatchNets Accuracy (20 classes): 20.5. Result ID: 6a1576dffbc373592c0d8c77. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Learning Setting":"Supervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. ### sDTW Accuracy (20 classes): 20.1. Result ID: 6a1576dffbc373592c0d8c7f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Learning Setting":"Unsupervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. ### Euclidean Accuracy (20 classes): 18.7. Result ID: 6a1576dffbc373592c0d8c7a. [Reporting paper]() Recorded evaluation configuration: {"Learning Setting":"Supervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. ### Euclidean Accuracy (20 classes): 13.5. Result ID: 6a1576dffbc373592c0d8c7e. [Reporting paper]() Recorded evaluation configuration: {"Learning Setting":"Unsupervised","Evaluation Protocol":"One-shot"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.