# Image Classification on SICAP C v2

Source: [Image Classification on SICAP C v2](<https://www.sota2.com/research/sota/image-classification-on-sicap-c-v2>)

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

Data updated: 2026-04-20T08:35:59.522Z

Evaluation of Image Classification performance on the SICAP C v2 dataset, reporting Semantic Consistency (PAC), Prototype Alignment (PCA-W), Prediction Stability, Selection Rate, and Relative Speed.

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) | SICAP C v2 |
| Metric | Semantic Consistency (PAC) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 7 of 7 |
| 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

- [Semantic Consistency (PAC)](<https://www.sota2.com/research/sota/image-classification-on-sicap-c-v2.md?metric=Semantic%20Consistency%20(PAC)>): Higher is better; unit: PERCENT.

- [Prototype Alignment (PCA-W)](<https://www.sota2.com/research/sota/image-classification-on-sicap-c-v2.md?metric=Prototype%20Alignment%20(PCA-W)>): Higher is better; unit: PERCENT.

- [Prediction Stability](<https://www.sota2.com/research/sota/image-classification-on-sicap-c-v2.md?metric=Prediction%20Stability>): Higher is better; unit: PERCENT.

- [Selection Rate](<https://www.sota2.com/research/sota/image-classification-on-sicap-c-v2.md?metric=Selection%20Rate>): Higher is better; unit: PERCENT.

- [Relative Speed](<https://www.sota2.com/research/sota/image-classification-on-sicap-c-v2.md?metric=Relative%20Speed>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Semantic Consistency (PAC) | Result date | Report paper |
| --- | --- | --- | --- |
| Unadapted | 88.8 | Not recorded | 69e5b30a7d0172140ec7df99 |
| ProtoTTA | 87.7 | Not recorded | 69e5b30a7d0172140ec7df99 |
| ProtoTTA+ | 87.4 | Not recorded | 69e5b30a7d0172140ec7df99 |
| EATA | 86.2 | Not recorded | 69e5b30a7d0172140ec7df99 |
| Memo | 86.1 | Not recorded | 69e5b30a7d0172140ec7df99 |
| SAR | 85.9 | Not recorded | 69e5b30a7d0172140ec7df99 |
| Tent | 84.5 | Not recorded | 69e5b30a7d0172140ec7df99 |

### Unadapted

Semantic Consistency (PAC): 88.8. Result ID: 69e5e56c47f34ce924cc1aa7.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b30a7d0172140ec7df99.md>)

Recorded evaluation configuration: {"Backbone":"ProtoPNet (BN + 1×1 Convs)"}. This record does not assert independent reproduction.

### ProtoTTA

Semantic Consistency (PAC): 87.7. Result ID: 69e5e56c47f34ce924cc1aac.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b30a7d0172140ec7df99.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e5b30a7d0172140ec7df99.md>)

Recorded evaluation configuration: {"Backbone":"ProtoPNet (BN + 1×1 Convs)"}. This record does not assert independent reproduction.

### ProtoTTA+

Semantic Consistency (PAC): 87.4. Result ID: 69e5e56c47f34ce924cc1aad.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b30a7d0172140ec7df99.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e5b30a7d0172140ec7df99.md>)

Recorded evaluation configuration: {"Backbone":"ProtoPNet (BN + 1×1 Convs)"}. This record does not assert independent reproduction.

### EATA

Semantic Consistency (PAC): 86.2. Result ID: 69e5e56c47f34ce924cc1aab.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b30a7d0172140ec7df99.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ca11e209131cea35123.md>)

Recorded evaluation configuration: {"Backbone":"ProtoPNet (BN + 1×1 Convs)"}. This record does not assert independent reproduction.

### Memo

Semantic Consistency (PAC): 86.1. Result ID: 69e5e56c47f34ce924cc1aa8.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b30a7d0172140ec7df99.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6da51e209131cea3b236.md>)

Recorded evaluation configuration: {"Backbone":"ProtoPNet (BN + 1×1 Convs)"}. This record does not assert independent reproduction.

### SAR

Semantic Consistency (PAC): 85.9. Result ID: 69e5e56c47f34ce924cc1aa9.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b30a7d0172140ec7df99.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e961e209131cea40e77.md>)

Recorded evaluation configuration: {"Backbone":"ProtoPNet (BN + 1×1 Convs)"}. This record does not assert independent reproduction.

### Tent

Semantic Consistency (PAC): 84.5. Result ID: 69e5e56c47f34ce924cc1aaa.

[Reporting paper](<https://www.sota2.com/research/paper/69e5b30a7d0172140ec7df99.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69941e209131cea22d2b.md>)

Recorded evaluation configuration: {"Backbone":"ProtoPNet (BN + 1×1 Convs)"}. This record does not assert independent reproduction.

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