# Semantic Segmentation on HSI-Drive 6-class experiment: pedestrians v2.0 v2.1

Source: [Semantic Segmentation on HSI-Drive 6-class experiment: pedestrians v2.0 v2.1](<https://www.sota2.com/research/sota/semantic-segmentation-on-hsi-drive-6-class-experiment-pedestrians-v2-0-v2-1>)

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

Data updated: 2026-03-27T13:28:18.052Z

Evaluation of Semantic Segmentation performance on the HSI-Drive 6-class experiment (versions v2.0 and v2.1), reporting per-class scores for road, vegetation, pedestrian, sky, others, plus global and weighted averages.

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 | Semantic Segmentation |
| Dataset (including any stated split) | HSI-Drive 6-class experiment: pedestrians v2.0 v2.1 |
| Metric | Road Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| 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

- [Road Accuracy](<https://www.sota2.com/research/sota/semantic-segmentation-on-hsi-drive-6-class-experiment-pedestrians-v2-0-v2-1.md?metric=Road%20Accuracy>): Higher is better; unit: PERCENT.

- [Road m. Score](<https://www.sota2.com/research/sota/semantic-segmentation-on-hsi-drive-6-class-experiment-pedestrians-v2-0-v2-1.md?metric=Road%20m.%20Score>): Higher is better; unit: PERCENT.

- [Vegetation Accuracy](<https://www.sota2.com/research/sota/semantic-segmentation-on-hsi-drive-6-class-experiment-pedestrians-v2-0-v2-1.md?metric=Vegetation%20Accuracy>): Higher is better; unit: PERCENT.

- [Pedestrian Accuracy](<https://www.sota2.com/research/sota/semantic-segmentation-on-hsi-drive-6-class-experiment-pedestrians-v2-0-v2-1.md?metric=Pedestrian%20Accuracy>): Higher is better; unit: PERCENT.

- [Sky Accuracy](<https://www.sota2.com/research/sota/semantic-segmentation-on-hsi-drive-6-class-experiment-pedestrians-v2-0-v2-1.md?metric=Sky%20Accuracy>): Higher is better; unit: PERCENT.

- [Others Accuracy](<https://www.sota2.com/research/sota/semantic-segmentation-on-hsi-drive-6-class-experiment-pedestrians-v2-0-v2-1.md?metric=Others%20Accuracy>): Higher is better; unit: PERCENT.

- [Global Accuracy](<https://www.sota2.com/research/sota/semantic-segmentation-on-hsi-drive-6-class-experiment-pedestrians-v2-0-v2-1.md?metric=Global%20Accuracy>): Higher is better; unit: PERCENT.

- [Weighted Accuracy](<https://www.sota2.com/research/sota/semantic-segmentation-on-hsi-drive-6-class-experiment-pedestrians-v2-0-v2-1.md?metric=Weighted%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Road Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Att.U-Net v2.1 | 97.6 | Not recorded | 69c60f370636e5bf4afc86be |
| U-Net v2.0 | 97.04 | Not recorded | 69c60f370636e5bf4afc86be |

### Att.U-Net v2.1

Road Accuracy: 97.6. Result ID: 69c685de47f34ce924b0a6e8.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f370636e5bf4afc86be.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c60f370636e5bf4afc86be.md>)

Recorded evaluation configuration: {"Version":"Scaling"}. This record does not assert independent reproduction.

### U-Net v2.0

Road Accuracy: 97.04. Result ID: 69c685de47f34ce924b0a6e7.

[Reporting paper](<https://www.sota2.com/research/paper/69c60f370636e5bf4afc86be.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75e41e209131cea6a80c.md>)

Recorded evaluation configuration: {"Version":"No scaling+PN"}. This record does not assert independent reproduction.

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