# Image Classification on DTD (Accuracy Gain, Epoch Gain)

Source: [Image Classification on DTD (Accuracy Gain, Epoch Gain)](<https://www.sota2.com/research/sota/image-classification-on-dtd-accuracy-gain-epoch-gain>)

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

Data updated: 2026-03-24T10:34:34.484Z

Evaluation of Image Classification performance on the Describable Textures Dataset (DTD), focusing on performance improvement metrics like Accuracy Gain and Epoch Gain.

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 | Image Classification |
| Dataset (including any stated split) | DTD |
| Metric | Accuracy Gain |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| 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 Gain](<https://www.sota2.com/research/sota/image-classification-on-dtd-accuracy-gain-epoch-gain.md?metric=Accuracy%20Gain>): Higher is better; unit: SCALAR.

- [Epoch Gain](<https://www.sota2.com/research/sota/image-classification-on-dtd-accuracy-gain-epoch-gain.md?metric=Epoch%20Gain>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Accuracy Gain | Result date | Report paper |
| --- | --- | --- | --- |
| targeted weight re-initialization | 2.4 | Not recorded | 69c21a6f0636e5bf4afc4700 |
| targeted weight re-initialization | 0.3 | Not recorded | 69c21a6f0636e5bf4afc4700 |
| targeted weight re-initialization | 0.15 | Not recorded | 69c21a6f0636e5bf4afc4700 |

### targeted weight re-initialization

Accuracy Gain: 2.4. Result ID: 69c2685247f34ce924aa5ca6.

[Reporting paper](<https://www.sota2.com/research/paper/69c21a6f0636e5bf4afc4700.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c21a6f0636e5bf4afc4700.md>)

Recorded evaluation configuration: {"Backbone":"SimpleNet"}. This record does not assert independent reproduction.

### targeted weight re-initialization

Accuracy Gain: 0.3. Result ID: 69c2685247f34ce924aa5ca7.

[Reporting paper](<https://www.sota2.com/research/paper/69c21a6f0636e5bf4afc4700.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c21a6f0636e5bf4afc4700.md>)

Recorded evaluation configuration: {"Backbone":"ResNet-50"}. This record does not assert independent reproduction.

### targeted weight re-initialization

Accuracy Gain: 0.15. Result ID: 69c2685247f34ce924aa5ca8.

[Reporting paper](<https://www.sota2.com/research/paper/69c21a6f0636e5bf4afc4700.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c21a6f0636e5bf4afc4700.md>)

Recorded evaluation configuration: {"Backbone":"ViT-B16"}. This record does not assert independent reproduction.

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