# Stereo Matching on Middlebury 2014 (test)

Source: [Stereo Matching on Middlebury 2014 (test)](<https://www.sota2.com/research/sota/stereo-matching-on-middlebury-2014-test>)

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

Data updated: 2026-02-26T05:37:29.137Z

Evaluation of Stereo Matching performance on the test split of the Middlebury 2014 dataset, reporting key error statistics such as Bad Pixel Rates (1.0 and 2.0 thresholds), Average Error, RMS Error, and 95th percentile error.

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 | Stereo Matching |
| Dataset (including any stated split) | Middlebury 2014 (test) |
| Metric | Bad Pixel Rate (Thresh 2.0) |
| Unit | PERCENT |
| Direction | Lower is better |
| Results loaded | 11 of 11 |
| 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

- [Bad Pixel Rate (Thresh 2.0)](<https://www.sota2.com/research/sota/stereo-matching-on-middlebury-2014-test.md?metric=Bad%20Pixel%20Rate%20(Thresh%202.0)>): Lower is better; unit: PERCENT.

- [Bad Pixel Rate (Thresh 1.0)](<https://www.sota2.com/research/sota/stereo-matching-on-middlebury-2014-test.md?metric=Bad%20Pixel%20Rate%20(Thresh%201.0)>): Lower is better; unit: PERCENT.

- [Average Error](<https://www.sota2.com/research/sota/stereo-matching-on-middlebury-2014-test.md?metric=Average%20Error>): Lower is better; unit: SCALAR.

- [RMS Error](<https://www.sota2.com/research/sota/stereo-matching-on-middlebury-2014-test.md?metric=RMS%20Error>): Lower is better; unit: SCALAR.

- [Error (95th Percentile)](<https://www.sota2.com/research/sota/stereo-matching-on-middlebury-2014-test.md?metric=Error%20(95th%20Percentile)>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Bad Pixel Rate (Thresh 2.0) | Result date | Report paper |
| --- | --- | --- | --- |
| CREStereo | 3.71 | Not recorded | 68fa6c0d1e209131cea31b3f |
| RAFT-Stereo | 4.74 | Not recorded | 68fa6c0d1e209131cea31b3f |
| LocalExp | 5.43 | Not recorded | 68fa6c0d1e209131cea31b3f |
| HITNet | 6.46 | Not recorded | 68fa6c0d1e209131cea31b3f |
| LEAStereo | 7.15 | Not recorded | 68fa6c0d1e209131cea31b3f |
| SDR | 7.69 | Not recorded | 68fa6c0d1e209131cea31b3f |
| MC-CNN-acrt | 8.08 | Not recorded | 68fa6c0d1e209131cea31b3f |
| CFNet | 10.1 | Not recorded | 68fa6c0d1e209131cea31b3f |
| HSMNet | 10.2 | Not recorded | 68fa6c0d1e209131cea31b3f |
| AdaStereo | 13.7 | Not recorded | 68fa6c0d1e209131cea31b3f |
| AANet++ | 15.4 | Not recorded | 68fa6c0d1e209131cea31b3f |

### CREStereo

Bad Pixel Rate (Thresh 2.0): 3.71. Result ID: 698ebd84ffc3901de0a03f13.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Implementation](<https://github.com/megvii-research/CREStereo>)

Recorded evaluation configuration: {"Resolution":"1536 x 2048","Inference mode":"2-stage"}. This record does not assert independent reproduction.

### RAFT-Stereo

Bad Pixel Rate (Thresh 2.0): 4.74. Result ID: 698ebd84ffc3901de0a03f14.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ad31e209131cea2a964.md>)

[Implementation](<https://github.com/princeton-vl/RAFT-Stereo>)

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

### LocalExp

Bad Pixel Rate (Thresh 2.0): 5.43. Result ID: 698ebd84ffc3901de0a03f15.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64e51e209131cea07386.md>)

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

### HITNet

Bad Pixel Rate (Thresh 2.0): 6.46. Result ID: 698ebd84ffc3901de0a03f16.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e661e209131cea3fb71.md>)

[Implementation](<https://github.com/google-research/google-research/tree/master/hitnet>)

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

### LEAStereo

Bad Pixel Rate (Thresh 2.0): 7.15. Result ID: 698ebd84ffc3901de0a03f17.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa689b1e209131cea1cfdf.md>)

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

### SDR

Bad Pixel Rate (Thresh 2.0): 7.69. Result ID: 698ebd84ffc3901de0a03f18.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

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

### MC-CNN-acrt

Bad Pixel Rate (Thresh 2.0): 8.08. Result ID: 698ebd84ffc3901de0a03f19.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa622c1e209131ce9f7dda.md>)

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

### CFNet

Bad Pixel Rate (Thresh 2.0): 10.1. Result ID: 698ebd84ffc3901de0a03f1a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69b81e209131cea23b1c.md>)

[Implementation](<https://github.com/gallenszl/CFNet>)

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

### HSMNet

Bad Pixel Rate (Thresh 2.0): 10.2. Result ID: 698ebd84ffc3901de0a03f1b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68391e209131cea1ae78.md>)

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

### AdaStereo

Bad Pixel Rate (Thresh 2.0): 13.7. Result ID: 698ebd84ffc3901de0a03f1c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa699e1e209131cea23207.md>)

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

### AANet++

Bad Pixel Rate (Thresh 2.0): 15.4. Result ID: 698ebd84ffc3901de0a03f1d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31b3f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa675b1e209131cea15d31.md>)

[Implementation](<https://github.com/haofeixu/aanet>)

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.
