# Multi-label image classification on VG-500 (test)

Source: [Multi-label image classification on VG-500 (test)](<https://www.sota2.com/research/sota/multi-label-image-classification-on-vg-500-test>)

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

Data updated: 2026-02-26T06:09:28.298Z

Evaluation of Multi-label Image Classification performance on the test split of the VG-500 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 | Multi-label image classification |
| Dataset (including any stated split) | VG-500 (test) |
| Metric | mAP |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| 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

- [mAP](<https://www.sota2.com/research/sota/multi-label-image-classification-on-vg-500-test.md?metric=mAP>): Higher is better; unit: PERCENT.

## Recorded results

| Method | mAP | Result date | Report paper |
| --- | --- | --- | --- |
| ML-Decoder + AAM | 43.1 | Not recorded | 68fa6e441e209131cea3eda1 |
| Q2L | 42.5 | Not recorded | 68fa6e441e209131cea3eda1 |
| ML-Decoder + AAM | 42 | Not recorded | 68fa6e441e209131cea3eda1 |
| ML-Decoder | 41.2 | Not recorded | 68fa6e441e209131cea3eda1 |
| Q2L | 40.35 | Not recorded | 68fa6e441e209131cea3eda1 |
| ASL | 38.84 | Not recorded | 68fa6e441e209131cea3eda1 |
| C-Tran | 38.4 | Not recorded | 68fa6e441e209131cea3eda1 |
| SSGRL | 36.6 | Not recorded | 68fa66371e209131cea0eda5 |
| ResNet-SRN | 33.5 | Not recorded | 68fa66371e209131cea0eda5 |
| ResNet-101 | 30.9 | Not recorded | 68fa66371e209131cea0eda5 |

### ML-Decoder + AAM

mAP: 43.1. Result ID: 69900b083a6c314849fdb5b2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e441e209131cea3eda1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e441e209131cea3eda1.md>)

[Implementation](<https://github.com/openvinotoolkit/deep-object-reid/tree/multilabel>)

Recorded evaluation configuration: {"Backbone":"TResNet-L","Input resolution":"576x576","GFLOPS":59.63}. This record does not assert independent reproduction.

### Q2L

mAP: 42.5. Result ID: 69900b083a6c314849fdb5ad.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e441e209131cea3eda1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a731e209131cea28343.md>)

[Implementation](<https://github.com/SlongLiu/query2labels>)

Recorded evaluation configuration: {"Backbone":"TResNet-L","Input resolution":"512x512","GFLOPS":119.37}. This record does not assert independent reproduction.

### ML-Decoder + AAM

mAP: 42. Result ID: 69900b083a6c314849fdb5b1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e441e209131cea3eda1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e441e209131cea3eda1.md>)

[Implementation](<https://github.com/openvinotoolkit/deep-object-reid/tree/multilabel>)

Recorded evaluation configuration: {"Backbone":"EfficientNet-V2-s","Input resolution":"576x576","GFLOPS":20.16}. This record does not assert independent reproduction.

### ML-Decoder

mAP: 41.2. Result ID: 69900b083a6c314849fdb5b0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e441e209131cea3eda1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b861e209131cea2ed73.md>)

[Implementation](<https://github.com/Alibaba-MIIL/ML_Decoder>)

Recorded evaluation configuration: {"Backbone":"EfficientNet-V2-s","Input resolution":"576x576","GFLOPS":20.16,"Training strategy":"Trained by current paper"}. This record does not assert independent reproduction.

### Q2L

mAP: 40.35. Result ID: 69900b083a6c314849fdb5af.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e441e209131cea3eda1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a731e209131cea28343.md>)

[Implementation](<https://github.com/SlongLiu/query2labels>)

Recorded evaluation configuration: {"Backbone":"EfficientNet-V2-s","Input resolution":"576x576","GFLOPS":32.81,"Training strategy":"Trained by current paper"}. This record does not assert independent reproduction.

### ASL

mAP: 38.84. Result ID: 69900b083a6c314849fdb5ae.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e441e209131cea3eda1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a6f1e209131cea280b7.md>)

[Implementation](<https://github.com/Alibaba-MIIL/ASL>)

Recorded evaluation configuration: {"Backbone":"EfficientNet-V2-s","Input resolution":"576x576","GFLOPS":17.9,"Training strategy":"Trained by current paper"}. This record does not assert independent reproduction.

### C-Tran

mAP: 38.4. Result ID: 69900b083a6c314849fdb5ac.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6e441e209131cea3eda1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68da1e209131cea1eb01.md>)

Recorded evaluation configuration: {"Backbone":"ResNet101","Input resolution":"576x576"}. This record does not assert independent reproduction.

### SSGRL

mAP: 36.6. Result ID: 698ce1298b517eb4331d63a2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66371e209131cea0eda5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66371e209131cea0eda5.md>)

[Implementation](<https://github.com/HCPLab-SYSU/SSGRL>)

Recorded evaluation configuration: {"backbone":"ResNet-101"}. This record does not assert independent reproduction.

### ResNet-SRN

mAP: 33.5. Result ID: 698ce1298b517eb4331d63a1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66371e209131cea0eda5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa633a1e209131ce9fda6c.md>)

[Implementation](<https://github.com/zhufengx/SRN_multilabel.>)

Recorded evaluation configuration: {"backbone":"ResNet-101"}. This record does not assert independent reproduction.

### ResNet-101

mAP: 30.9. Result ID: 698ce1298b517eb4331d63a0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa66371e209131cea0eda5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62321e209131ce9f8328.md>)

Recorded evaluation configuration: {"backbone":"ResNet-101"}. This record does not assert independent reproduction.

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