# Skin Lesion Classification on ISIC 3-way 3-shot 2018 (test) Source: [Skin Lesion Classification on ISIC 3-way 3-shot 2018 (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:28:01.065Z Evaluation of skin lesion classification performance using the 3-way, 3-shot setup on the test split of the ISIC 2018 dataset. 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 | Skin Lesion Classification | | Dataset (including any stated split) | ISIC 3-way 3-shot 2018 (test) | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 28 of 28 | | 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. - [F1 Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Hybrid distil | 75.82 | Not recorded | 68fa6afd1e209131cea2b941 | | MoCo\_v1 (baseline) | 74.92 | Not recorded | 68fa6afd1e209131cea2b941 | | MoCo\_v2 | 70.84 | Not recorded | 68fa6afd1e209131cea2b941 | | Hybrid distil | 69.14 | Not recorded | 68fa6afd1e209131cea2b941 | | MoCo\_v1 (baseline) | 67.48 | Not recorded | 68fa6afd1e209131cea2b941 | | SRL-distil | 64.92 | Not recorded | 68fa6afd1e209131cea2b941 | | SimCLR | 63.82 | Not recorded | 68fa6afd1e209131cea2b941 | | SRL-simple | 61.31 | Not recorded | 68fa6afd1e209131cea2b941 | | SimCLR | 60.78 | Not recorded | 68fa6afd1e209131cea2b941 | | MoCo\_v2 | 59.71 | Not recorded | 68fa6afd1e209131cea2b941 | | SRL-distil | 58.2 | Not recorded | 68fa6afd1e209131cea2b941 | | SRL-simple | 57.5 | Not recorded | 68fa6afd1e209131cea2b941 | | DAML | 55.57 | Not recorded | 68fa6afd1e209131cea2b941 | | MAML | 55.55 | Not recorded | 68fa6afd1e209131cea2b941 | | DAML | 53.91 | Not recorded | 68fa6afd1e209131cea2b941 | | MAML | 51.56 | Not recorded | 68fa6afd1e209131cea2b941 | | UMTRA | 51.29 | Not recorded | 68fa6afd1e209131cea2b941 | | RelationNet | 47.29 | Not recorded | 68fa6afd1e209131cea2b941 | | UMTRA | 45.76 | Not recorded | 68fa6afd1e209131cea2b941 | | CACTUS-ProtoNets | 45 | Not recorded | 68fa6afd1e209131cea2b941 | | CACTUS-MAML | 44.44 | Not recorded | 68fa6afd1e209131cea2b941 | | RelationNet | 42.63 | Not recorded | 68fa6afd1e209131cea2b941 | | CACTUS-MAML | 42.2 | Not recorded | 68fa6afd1e209131cea2b941 | | Training from scratch | 39.76 | Not recorded | 68fa6afd1e209131cea2b941 | | Training from scratch | 38.95 | Not recorded | 68fa6afd1e209131cea2b941 | | ProtoNets | 38.59 | Not recorded | 68fa6afd1e209131cea2b941 | | ProtoNets | 36.38 | Not recorded | 68fa6afd1e209131cea2b941 | | CACTUS-ProtoNets | 35.08 | Not recorded | 68fa6afd1e209131cea2b941 | ### Hybrid distil Accuracy: 75.82. Result ID: 698e4ac0ffc3901de09c5ee6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### MoCo\_v1 (baseline) Accuracy: 74.92. Result ID: 698e4ac0ffc3901de09c5ee5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### MoCo\_v2 Accuracy: 70.84. Result ID: 698e4ac0ffc3901de09c5ee4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### Hybrid distil Accuracy: 69.14. Result ID: 698e4ac0ffc3901de09c5ebc. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### MoCo\_v1 (baseline) Accuracy: 67.48. Result ID: 698e4ac0ffc3901de09c5ebb. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### SRL-distil Accuracy: 64.92. Result ID: 698e4ac0ffc3901de09c5ee2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### SimCLR Accuracy: 63.82. Result ID: 698e4ac0ffc3901de09c5ee3. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### SRL-simple Accuracy: 61.31. Result ID: 698e4ac0ffc3901de09c5ee1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### SimCLR Accuracy: 60.78. Result ID: 698e4ac0ffc3901de09c5eb9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### MoCo\_v2 Accuracy: 59.71. Result ID: 698e4ac0ffc3901de09c5eba. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### SRL-distil Accuracy: 58.2. Result ID: 698e4ac0ffc3901de09c5eb8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### SRL-simple Accuracy: 57.5. Result ID: 698e4ac0ffc3901de09c5eb7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### DAML Accuracy: 55.57. Result ID: 698e4ac0ffc3901de09c5edd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### MAML Accuracy: 55.55. Result ID: 698e4ac0ffc3901de09c5eda. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### DAML Accuracy: 53.91. Result ID: 698e4ac0ffc3901de09c5eb3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### MAML Accuracy: 51.56. Result ID: 698e4ac0ffc3901de09c5eb0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### UMTRA Accuracy: 51.29. Result ID: 698e4ac0ffc3901de09c5ede. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### RelationNet Accuracy: 47.29. Result ID: 698e4ac0ffc3901de09c5edb. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### UMTRA Accuracy: 45.76. Result ID: 698e4ac0ffc3901de09c5eb4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### CACTUS-ProtoNets Accuracy: 45. Result ID: 698e4ac0ffc3901de09c5ee0. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### CACTUS-MAML Accuracy: 44.44. Result ID: 698e4ac0ffc3901de09c5edf. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### RelationNet Accuracy: 42.63. Result ID: 698e4ac0ffc3901de09c5eb1. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### CACTUS-MAML Accuracy: 42.2. Result ID: 698e4ac0ffc3901de09c5eb5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### Training from scratch Accuracy: 39.76. Result ID: 698e4ac0ffc3901de09c5ed9. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### Training from scratch Accuracy: 38.95. Result ID: 698e4ac0ffc3901de09c5eaf. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### ProtoNets Accuracy: 38.59. Result ID: 698e4ac0ffc3901de09c5edc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-12"}. This record does not assert independent reproduction. ### ProtoNets Accuracy: 36.38. Result ID: 698e4ac0ffc3901de09c5eb2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. ### CACTUS-ProtoNets Accuracy: 35.08. Result ID: 698e4ac0ffc3901de09c5eb6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"4 conv blocks"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.