# Object-Counting on Extreme Illumination Triplet Instruct-Following Dataset

Source: [Object-Counting on Extreme Illumination Triplet Instruct-Following Dataset](<https://www.sota2.com/research/sota/object-counting-on-extreme-illumination-triplet-instruct-following-dataset>)

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

Data updated: 2026-03-31T09:53:32.061Z

Evaluation of model performance on the Extreme Illumination Triplet Instruct-Following Dataset for object counting, reporting Accuracy and Mean Absolute Error (MAE).

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 | Object-Counting |
| Dataset (including any stated split) | Extreme Illumination Triplet Instruct-Following Dataset |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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](<https://www.sota2.com/research/sota/object-counting-on-extreme-illumination-triplet-instruct-following-dataset.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [MAE](<https://www.sota2.com/research/sota/object-counting-on-extreme-illumination-triplet-instruct-following-dataset.md?metric=MAE>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Ours-7B | 74.66 | Not recorded | 69cb555f0636e5bf4afca24b |
| Baseline-7B | 72.73 | Not recorded | 69cb555f0636e5bf4afca24b |
| InternVL | 72.56 | Not recorded | 69cb555f0636e5bf4afca24b |
| Ours-3B | 71.08 | Not recorded | 69cb555f0636e5bf4afca24b |
| Baseline-3B | 69.66 | Not recorded | 69cb555f0636e5bf4afca24b |
| LLaVA-7B | 67.84 | Not recorded | 69cb555f0636e5bf4afca24b |
| EventGPT | 67.41 | Not recorded | 69cb555f0636e5bf4afca24b |
| Q-Instruct | 66.87 | Not recorded | 69cb555f0636e5bf4afca24b |
| Deepseek-VL | 54.87 | Not recorded | 69cb555f0636e5bf4afca24b |

### Ours-7B

Accuracy: 74.66. Result ID: 69cb998f47f34ce924b45725.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

Recorded evaluation configuration: {"MLLM Type":"Event-Enhanced Model","Backbone":"Qwen-7B"}. This record does not assert independent reproduction.

### Baseline-7B

Accuracy: 72.73. Result ID: 69cb998f47f34ce924b45724.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75151e209131cea65976.md>)

[Implementation](<https://github.com/QwenLM/Qwen2-VL>)

Recorded evaluation configuration: {"MLLM Type":"Event-Enhanced Model","Backbone":"Qwen-7B"}. This record does not assert independent reproduction.

### InternVL

Accuracy: 72.56. Result ID: 69cb998f47f34ce924b4571f.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78241e209131cea76fa5.md>)

Recorded evaluation configuration: {"MLLM Type":"General Model"}. This record does not assert independent reproduction.

### Ours-3B

Accuracy: 71.08. Result ID: 69cb998f47f34ce924b45723.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

Recorded evaluation configuration: {"MLLM Type":"Event-Enhanced Model","Backbone":"Qwen-3B"}. This record does not assert independent reproduction.

### Baseline-3B

Accuracy: 69.66. Result ID: 69cb998f47f34ce924b45722.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75151e209131cea65976.md>)

[Implementation](<https://github.com/QwenLM/Qwen2-VL>)

Recorded evaluation configuration: {"MLLM Type":"Event-Enhanced Model","Backbone":"Qwen-3B"}. This record does not assert independent reproduction.

### LLaVA-7B

Accuracy: 67.84. Result ID: 69cb998f47f34ce924b4571d.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa717d1e209131cea51970.md>)

[Implementation](<https://llava-vl.github.io/>)

Recorded evaluation configuration: {"MLLM Type":"General Model"}. This record does not assert independent reproduction.

### EventGPT

Accuracy: 67.41. Result ID: 69cb998f47f34ce924b45720.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76051e209131cea6af75.md>)

Recorded evaluation configuration: {"MLLM Type":"Event-specialized Model"}. This record does not assert independent reproduction.

### Q-Instruct

Accuracy: 66.87. Result ID: 69cb998f47f34ce924b45721.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa712c1e209131cea4fb99.md>)

Recorded evaluation configuration: {"MLLM Type":"Illumination-aware Model"}. This record does not assert independent reproduction.

### Deepseek-VL

Accuracy: 54.87. Result ID: 69cb998f47f34ce924b4571e.

[Reporting paper](<https://www.sota2.com/research/paper/69cb555f0636e5bf4afca24b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa763a1e209131cea6bfb0.md>)

Recorded evaluation configuration: {"MLLM Type":"General Model"}. This record does not assert independent reproduction.

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