# Counting Visual Question Answering on VQA Count-QA

Source: [Counting Visual Question Answering on VQA Count-QA](<https://www.sota2.com/research/sota/counting-visual-question-answering-on-vqa-count-qa>)

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

Data updated: 2026-02-26T06:11:02.447Z

Evaluation of model performance on the VQA Count-QA dataset specifically for counting visual questions, typically measured using mean Root Mean Square Error (mRMSE).

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 | Counting Visual Question Answering |
| Dataset (including any stated split) | VQA Count-QA |
| Metric | mRMSE |
| Unit | SCALAR |
| Direction | Lower 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

- [mRMSE](<https://www.sota2.com/research/sota/counting-visual-question-answering-on-vqa-count-qa.md?metric=mRMSE>): Lower is better; unit: SCALAR.

## Recorded results

| Method | mRMSE | Result date | Report paper |
| --- | --- | --- | --- |
| ens | 1.8 | Not recorded | 68fa63531e209131ce9fe2ae |
| seq-sub-ft-3x3 | 1.81 | Not recorded | 68fa63531e209131ce9fe2ae |
| aso-sub-ft-1L-3x3 | 1.94 | Not recorded | 68fa63531e209131ce9fe2ae |
| glance-ft-1L | 2.19 | Not recorded | 68fa63531e209131ce9fe2ae |
| Deeper LSTM | 2.71 | Not recorded | 68fa63531e209131ce9fe2ae |
| detect | 2.72 | Not recorded | 68fa63531e209131ce9fe2ae |
| SOTA VQA | 3.25 | Not recorded | 68fa63531e209131ce9fe2ae |

### ens

mRMSE: 1.8. Result ID: 698bf157ecace47ef55a869e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

Recorded evaluation configuration: {"Description":"Combination of glance-ft-1L, aso-sub-ft-1L-3x3 and seq-sub-ft-3x3"}. This record does not assert independent reproduction.

### seq-sub-ft-3x3

mRMSE: 1.81. Result ID: 698bf157ecace47ef55a869d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

Recorded evaluation configuration: {"Mode":"Fine-tuned","Grid":"3x3"}. This record does not assert independent reproduction.

### aso-sub-ft-1L-3x3

mRMSE: 1.94. Result ID: 698bf157ecace47ef55a869c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

Recorded evaluation configuration: {"Mode":"Fine-tuned","Hidden Layers":"1","Grid":"3x3"}. This record does not assert independent reproduction.

### glance-ft-1L

mRMSE: 2.19. Result ID: 698bf157ecace47ef55a869b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

Recorded evaluation configuration: {"Mode":"Fine-tuned","Hidden Layers":"1"}. This record does not assert independent reproduction.

### Deeper LSTM

mRMSE: 2.71. Result ID: 698bf157ecace47ef55a869f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### detect

mRMSE: 2.72. Result ID: 698bf157ecace47ef55a869a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62191e209131ce9f79f3.md>)

[Implementation](<https://github.com/rbgirshick/fast-rcnn>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### SOTA VQA

mRMSE: 3.25. Result ID: 698bf157ecace47ef55a86a0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63531e209131ce9fe2ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62bc1e209131ce9faf5e.md>)

[Implementation](<https://github.com/akirafukui/vqa-mcb>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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