# Incremental Learning on ImageNet-1000 10 steps 100 classes per step (test)

Source: [Incremental Learning on ImageNet-1000 10 steps 100 classes per step (test)](<https://www.sota2.com/research/sota/incremental-learning-on-imagenet-1000-10-steps-100-classes-per-step-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:35:55.241Z

Evaluation of Incremental Learning performance on a specific subset configuration of ImageNet-1000 (10 steps, 100 classes per step) evaluated on the test split, using average session accuracy.

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 | Incremental Learning |
| Dataset (including any stated split) | ImageNet-1000 10 steps 100 classes per step (test) |
| Metric | Average Session Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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

- [Average Session Accuracy](<https://www.sota2.com/research/sota/incremental-learning-on-imagenet-1000-10-steps-100-classes-per-step-test.md?metric=Average%20Session%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Average Session Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Bound | 89.27 | Not recorded | 68fa6d181e209131cea37af3 |
| FOSTER | 68.34 | Not recorded | 68fa6d181e209131cea37af3 |
| DER | 66.73 | Not recorded | 68fa6d181e209131cea37af3 |
| WA | 54.1 | Not recorded | 68fa6d181e209131cea37af3 |
| iCaRL | 38.4 | Not recorded | 68fa6d181e209131cea37af3 |

### Bound

Average Session Accuracy: 89.27. Result ID: 698f8db33a6c314849fa0cdd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d181e209131cea37af3.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### FOSTER

Average Session Accuracy: 68.34. Result ID: 698f8db33a6c314849fa0ce1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d181e209131cea37af3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d181e209131cea37af3.md>)

[Implementation](<https://github.com/G-U-N/ECCV22-FOSTER>)

Recorded evaluation configuration: {"backbones":"single"}. This record does not assert independent reproduction.

### DER

Average Session Accuracy: 66.73. Result ID: 698f8db33a6c314849fa0ce0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d181e209131cea37af3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa699b1e209131cea23084.md>)

[Implementation](<https://github.com/Rhyssiyan/DER-ClassIL.pytorch>)

Recorded evaluation configuration: {"backbones":"multiple (incremental backbones)"}. This record does not assert independent reproduction.

### WA

Average Session Accuracy: 54.1. Result ID: 698f8db33a6c314849fa0cdf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d181e209131cea37af3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66aa1e209131cea117b4.md>)

Recorded evaluation configuration: {"backbones":"single"}. This record does not assert independent reproduction.

### iCaRL

Average Session Accuracy: 38.4. Result ID: 698f8db33a6c314849fa0cde.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6d181e209131cea37af3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa634e1e209131ce9fded4.md>)

Recorded evaluation configuration: {"backbones":"single"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.