# AI-Generated Image Detection on GenImage v1.4 (test)

Source: [AI-Generated Image Detection on GenImage v1.4 (test)](<https://www.sota2.com/research/sota/ai-generated-image-detection-on-genimage-v1-4-test>)

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

Data updated: 2026-05-29T09:53:29.428Z

Evaluation of AI-Generated Image Detection performance on the test split of the GenImage v1.4 dataset, reporting detection rates/accuracy for specific generative models including Midjourney, Stable Diffusion variants, ADM, GLIDE, Wukong, VQDM, and BigGAN.

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 | AI-Generated Image Detection |
| Dataset (including any stated split) | GenImage v1.4 (test) |
| Metric | Midjourney Detection Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 15 of 15 |
| 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

- [Midjourney Detection Accuracy](<https://www.sota2.com/research/sota/ai-generated-image-detection-on-genimage-v1-4-test.md?metric=Midjourney%20Detection%20Accuracy>): Higher is better; unit: PERCENT.

- [SD v1.4 Detection Accuracy](<https://www.sota2.com/research/sota/ai-generated-image-detection-on-genimage-v1-4-test.md?metric=SD%20v1.4%20Detection%20Accuracy>): Higher is better; unit: PERCENT.

- [SD v1.5 Detection Accuracy](<https://www.sota2.com/research/sota/ai-generated-image-detection-on-genimage-v1-4-test.md?metric=SD%20v1.5%20Detection%20Accuracy>): Higher is better; unit: PERCENT.

- [ADM Detection Accuracy](<https://www.sota2.com/research/sota/ai-generated-image-detection-on-genimage-v1-4-test.md?metric=ADM%20Detection%20Accuracy>): Higher is better; unit: PERCENT.

- [GLIDE Detection Accuracy](<https://www.sota2.com/research/sota/ai-generated-image-detection-on-genimage-v1-4-test.md?metric=GLIDE%20Detection%20Accuracy>): Higher is better; unit: PERCENT.

- [Wukong Detection Accuracy](<https://www.sota2.com/research/sota/ai-generated-image-detection-on-genimage-v1-4-test.md?metric=Wukong%20Detection%20Accuracy>): Higher is better; unit: PERCENT.

- [VQDM Detection Accuracy](<https://www.sota2.com/research/sota/ai-generated-image-detection-on-genimage-v1-4-test.md?metric=VQDM%20Detection%20Accuracy>): Higher is better; unit: PERCENT.

- [BigGAN Detection Accuracy](<https://www.sota2.com/research/sota/ai-generated-image-detection-on-genimage-v1-4-test.md?metric=BigGAN%20Detection%20Accuracy>): Higher is better; unit: PERCENT.

- [Mean Detection Accuracy](<https://www.sota2.com/research/sota/ai-generated-image-detection-on-genimage-v1-4-test.md?metric=Mean%20Detection%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Midjourney Detection Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| OmniAID-Mirage | 98 | Not recorded | 6914148d2828b7cc6937ebaf |
| FatFormer | 92.7 | Not recorded | 6914148d2828b7cc6937ebaf |
| UnivFD | 91.5 | Not recorded | 6914148d2828b7cc6937ebaf |
| DRCT | 91.5 | Not recorded | 6914148d2828b7cc6937ebaf |
| GenDet | 89.6 | Not recorded | 6914148d2828b7cc6937ebaf |
| OmniAID | 85.7 | Not recorded | 6914148d2828b7cc6937ebaf |
| Effort | 82.4 | Not recorded | 6914148d2828b7cc6937ebaf |
| NPR | 81 | Not recorded | 6914148d2828b7cc6937ebaf |
| AIDE | 79.4 | Not recorded | 6914148d2828b7cc6937ebaf |
| PatchCraft | 79 | Not recorded | 6914148d2828b7cc6937ebaf |
| DIRE | 60.2 | Not recorded | 6914148d2828b7cc6937ebaf |
| GramNet | 54.2 | Not recorded | 6914148d2828b7cc6937ebaf |
| CNNSpot | 52.8 | Not recorded | 6914148d2828b7cc6937ebaf |
| Spec | 52 | Not recorded | 6914148d2828b7cc6937ebaf |
| F3Net | 50.1 | Not recorded | 6914148d2828b7cc6937ebaf |

### OmniAID-Mirage

Midjourney Detection Accuracy: 98. Result ID: 6a1960d3fbc373592c142770.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

Recorded evaluation configuration: {"Training Data":"Mirage-Train"}. This record does not assert independent reproduction.

### FatFormer

Midjourney Detection Accuracy: 92.7. Result ID: 6a1960d3fbc373592c14276b.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71a11e209131cea52333.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### UnivFD

Midjourney Detection Accuracy: 91.5. Result ID: 6a1960d3fbc373592c142767.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72b11e209131cea57ee9.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### DRCT

Midjourney Detection Accuracy: 91.5. Result ID: 6a1960d3fbc373592c14276c.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### GenDet

Midjourney Detection Accuracy: 89.6. Result ID: 6a1960d3fbc373592c142768.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71721e209131cea514c7.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### OmniAID

Midjourney Detection Accuracy: 85.7. Result ID: 6a1960d3fbc373592c14276f.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### Effort

Midjourney Detection Accuracy: 82.4. Result ID: 6a1960d3fbc373592c14276e.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a71e209131cea79a6c.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### NPR

Midjourney Detection Accuracy: 81. Result ID: 6a1960d3fbc373592c14276a.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71821e209131cea51af9.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### AIDE

Midjourney Detection Accuracy: 79.4. Result ID: 6a1960d3fbc373592c14276d.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77061e209131cea708f9.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### PatchCraft

Midjourney Detection Accuracy: 79. Result ID: 6a1960d3fbc373592c142769.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72781e209131cea56910.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### DIRE

Midjourney Detection Accuracy: 60.2. Result ID: 6a1960d3fbc373592c142766.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ec61e209131cea41f4c.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### GramNet

Midjourney Detection Accuracy: 54.2. Result ID: 6a1960d3fbc373592c142765.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa667d1e209131cea108e7.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### CNNSpot

Midjourney Detection Accuracy: 52.8. Result ID: 6a1960d3fbc373592c142762.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67581e209131cea15b26.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### Spec

Midjourney Detection Accuracy: 52. Result ID: 6a1960d3fbc373592c142763.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa667d1e209131cea108e7.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

### F3Net

Midjourney Detection Accuracy: 50.1. Result ID: 6a1960d3fbc373592c142764.

[Reporting paper](<https://www.sota2.com/research/paper/6914148d2828b7cc6937ebaf.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa689c1e209131cea1d080.md>)

Recorded evaluation configuration: {"Training Data":"GenImage-SD v1.4"}. This record does not assert independent reproduction.

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