# Affordance Grounding on UMD (test)

Source: [Affordance Grounding on UMD (test)](<https://www.sota2.com/research/sota/affordance-grounding-on-umd-test>)

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

Data updated: 2026-05-04T07:51:32.660Z

Evaluation of Affordance Grounding performance on the test split of the UMD dataset, measured using gIoU, cIoU, P50, and P50-95.

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 | Affordance Grounding |
| Dataset (including any stated split) | UMD (test) |
| Metric | gIoU |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 18 of 18 |
| 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

- [gIoU](<https://www.sota2.com/research/sota/affordance-grounding-on-umd-test.md?metric=gIoU>): Higher is better; unit: PERCENT.

- [cIoU](<https://www.sota2.com/research/sota/affordance-grounding-on-umd-test.md?metric=cIoU>): Higher is better; unit: PERCENT.

- [P50](<https://www.sota2.com/research/sota/affordance-grounding-on-umd-test.md?metric=P50>): Higher is better; unit: PERCENT.

- [P50-95](<https://www.sota2.com/research/sota/affordance-grounding-on-umd-test.md?metric=P50-95>): Higher is better; unit: PERCENT.

## Recorded results

| Method | gIoU | Result date | Report paper |
| --- | --- | --- | --- |
| A-Harness | 57.61 | Not recorded | 69f827e7c0f56ed8430888ed |
| A-Harness | 54.94 | Not recorded | 69f827e7c0f56ed8430888ed |
| A-Harness | 54.28 | Not recorded | 69f827e7c0f56ed8430888ed |
| A-Harness | 53.72 | Not recorded | 69f827e7c0f56ed8430888ed |
| A-Harness | 52.74 | Not recorded | 69f827e7c0f56ed8430888ed |
| A-Harness | 51.33 | Not recorded | 69f827e7c0f56ed8430888ed |
| Full Fixed Skill Chain | 50.19 | Not recorded | 69f827e7c0f56ed8430888ed |
| Affordance-R1 | 49.85 | Not recorded | 69f827e7c0f56ed8430888ed |
| Only w/ Det. & Seg. skills | 46.53 | Not recorded | 69f827e7c0f56ed8430888ed |
| Seg-Zero | 44.26 | Not recorded | 69f827e7c0f56ed8430888ed |
| Vision Reasoner | 44 | Not recorded | 69f827e7c0f56ed8430888ed |
| AffordanceLLM | 43.11 | Not recorded | 69f827e7c0f56ed8430888ed |
| LISA-7B | 41.9 | Not recorded | 69f827e7c0f56ed8430888ed |
| ConverSeg | 33.27 | Not recorded | 69f827e7c0f56ed8430888ed |
| Qwen2.5VL-7B | 33.21 | Not recorded | 69f827e7c0f56ed8430888ed |
| InternVL3-7B | 30.46 | Not recorded | 69f827e7c0f56ed8430888ed |
| AffordanceVLM | 25.41 | Not recorded | 69f827e7c0f56ed8430888ed |
| SAM4MLLM | 12.4 | Not recorded | 69f827e7c0f56ed8430888ed |

### A-Harness

gIoU: 57.61. Result ID: 69f84fef47f34ce924dbc45b.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

Recorded evaluation configuration: {"Harness Setup":"w/ A-Harness","Decision Brain":"Qwen-3.5-397B-A17B"}. This record does not assert independent reproduction.

### A-Harness

gIoU: 54.94. Result ID: 69f84fef47f34ce924dbc460.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

Recorded evaluation configuration: {"Harness Setup":"w/ A-Harness","Decision Brain":"Claude-Opus-4.6"}. This record does not assert independent reproduction.

### A-Harness

gIoU: 54.28. Result ID: 69f84fef47f34ce924dbc45e.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

Recorded evaluation configuration: {"Harness Setup":"w/ A-Harness","Decision Brain":"GLM-5"}. This record does not assert independent reproduction.

### A-Harness

gIoU: 53.72. Result ID: 69f84fef47f34ce924dbc45f.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

Recorded evaluation configuration: {"Harness Setup":"w/ A-Harness","Decision Brain":"Claude-Sonnet-4.6"}. This record does not assert independent reproduction.

### A-Harness

gIoU: 52.74. Result ID: 69f84fef47f34ce924dbc45d.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

Recorded evaluation configuration: {"Harness Setup":"w/ A-Harness","Decision Brain":"GPT-4o"}. This record does not assert independent reproduction.

### A-Harness

gIoU: 51.33. Result ID: 69f84fef47f34ce924dbc45c.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

Recorded evaluation configuration: {"Harness Setup":"w/ A-Harness","Decision Brain":"Gemini-3-flash"}. This record does not assert independent reproduction.

### Full Fixed Skill Chain

gIoU: 50.19. Result ID: 69f84fef47f34ce924dbc45a.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

Recorded evaluation configuration: {"Harness Setup":"w/o A-Harness"}. This record does not assert independent reproduction.

### Affordance-R1

gIoU: 49.85. Result ID: 69f84fef47f34ce924dbc457.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7aa51e209131cea845ac.md>)

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

### Only w/ Det. & Seg. skills

gIoU: 46.53. Result ID: 69f84fef47f34ce924dbc459.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

Recorded evaluation configuration: {"Harness Setup":"w/o A-Harness"}. This record does not assert independent reproduction.

### Seg-Zero

gIoU: 44.26. Result ID: 69f84fef47f34ce924dbc455.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

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

### Vision Reasoner

gIoU: 44. Result ID: 69f84fef47f34ce924dbc456.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ba21e209131cea897a6.md>)

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

### AffordanceLLM

gIoU: 43.11. Result ID: 69f84fef47f34ce924dbc451.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72eb1e209131cea59389.md>)

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

### LISA-7B

gIoU: 41.9. Result ID: 69f84fef47f34ce924dbc44f.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa730a1e209131cea5a0b2.md>)

[Implementation](<https://www.sota2.com/github.com/dvlab-research/LISA>)

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

### ConverSeg

gIoU: 33.27. Result ID: 69f84fef47f34ce924dbc458.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6992a4c39ebc499c97406e0f.md>)

[Implementation](<https://github.com/AadSah/ConverSeg>)

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

### Qwen2.5VL-7B

gIoU: 33.21. Result ID: 69f84fef47f34ce924dbc453.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

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

### InternVL3-7B

gIoU: 30.46. Result ID: 69f84fef47f34ce924dbc452.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

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

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

### AffordanceVLM

gIoU: 25.41. Result ID: 69f84fef47f34ce924dbc454.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a551e209131cea828ef.md>)

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

### SAM4MLLM

gIoU: 12.4. Result ID: 69f84fef47f34ce924dbc450.

[Reporting paper](<https://www.sota2.com/research/paper/69f827e7c0f56ed8430888ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76371e209131cea6bea9.md>)

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.
