# Referring Expression Comprehension on RefCOCO (Comprehensive Metrics)

Source: [Referring Expression Comprehension on RefCOCO (Comprehensive Metrics)](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics>)

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

Data updated: 2026-06-17T07:52:55.073Z

Evaluation of Referring Expression Comprehension performance on the RefCOCO dataset, reporting a comprehensive set of metrics including Instance/Deletion IoU, Point Game scores, and MEC scores.

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 | Referring Expression Comprehension |
| Dataset (including any stated split) | RefCOCO |
| Metric | Instance IoU |
| 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

- [Instance IoU](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=Instance%20IoU>): Higher is better; unit: PERCENT.

- [Deletion IoU](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=Deletion%20IoU>): Higher is better; unit: PERCENT.

- [Instance IoU (Class)](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=Instance%20IoU%20(Class)>): Higher is better; unit: PERCENT.

- [Deletion IoU (Class)](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=Deletion%20IoU%20(Class)>): Higher is better; unit: PERCENT.

- [Instance IoU Score](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=Instance%20IoU%20Score>): Higher is better; unit: PERCENT.

- [Deletion IoU Score](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=Deletion%20IoU%20Score>): Higher is better; unit: PERCENT.

- [Average High IoU Score](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=Average%20High%20IoU%20Score>): Higher is better; unit: PERCENT.

- [Point Game Score](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=Point%20Game%20Score>): Higher is better; unit: PERCENT.

- [Energy Point Game Score](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=Energy%20Point%20Game%20Score>): Higher is better; unit: PERCENT.

- [Average Minimum Error Count (MEC)](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=Average%20Minimum%20Error%20Count%20(MEC)>): Lower is better; unit: SCALAR.

- [A-C Ratio](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-comprehensive-metrics.md?metric=A-C%20Ratio>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Instance IoU | Result date | Report paper |
| --- | --- | --- | --- |
| Greedy | 74.19 | Not recorded | 6a322a57d33dfb23e783ed5d |
| PhaseWin | 73.77 | Not recorded | 6a322a57d33dfb23e783ed5d |
| Greedy | 72.78 | Not recorded | 6a322a57d33dfb23e783ed5d |
| PhaseWin | 70.13 | Not recorded | 6a322a57d33dfb23e783ed5d |
| RISE | 61.78 | Not recorded | 6a322a57d33dfb23e783ed5d |
| HSIC | 54.91 | Not recorded | 6a322a57d33dfb23e783ed5d |
| ODAM | 47.78 | Not recorded | 6a322a57d33dfb23e783ed5d |
| SSGrad-CAM++ | 41.13 | Not recorded | 6a322a57d33dfb23e783ed5d |
| Grad-CAM | 37.49 | Not recorded | 6a322a57d33dfb23e783ed5d |

### Greedy

Instance IoU: 74.19. Result ID: 6a325254fbc373592c2fa658.

[Reporting paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74c31e209131cea63a16.md>)

[Implementation](<https://github.com/RuoyuChen10/SMDL-Attribution>)

Recorded evaluation configuration: {"Model":"Grounding DINO","Number of regions":"100"}. This record does not assert independent reproduction.

### PhaseWin

Instance IoU: 73.77. Result ID: 6a325254fbc373592c2fa659.

[Reporting paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

Recorded evaluation configuration: {"Model":"Grounding DINO","Number of regions":"100"}. This record does not assert independent reproduction.

### Greedy

Instance IoU: 72.78. Result ID: 6a325254fbc373592c2fa656.

[Reporting paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74c31e209131cea63a16.md>)

[Implementation](<https://github.com/RuoyuChen10/SMDL-Attribution>)

Recorded evaluation configuration: {"Model":"Grounding DINO","Number of regions":"50"}. This record does not assert independent reproduction.

### PhaseWin

Instance IoU: 70.13. Result ID: 6a325254fbc373592c2fa657.

[Reporting paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

Recorded evaluation configuration: {"Model":"Grounding DINO","Number of regions":"50"}. This record does not assert independent reproduction.

### RISE

Instance IoU: 61.78. Result ID: 6a325254fbc373592c2fa653.

[Reporting paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64dc1e209131cea06cb9.md>)

[Implementation](<https://github.com/pytorch/vision>)

Recorded evaluation configuration: {"Model":"Grounding DINO"}. This record does not assert independent reproduction.

### HSIC

Instance IoU: 54.91. Result ID: 6a325254fbc373592c2fa654.

[Reporting paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d891e209131cea3a77b.md>)

[Implementation](<https://github.com/paulnovello/HSIC-Attribution-Method>)

Recorded evaluation configuration: {"Model":"Grounding DINO"}. This record does not assert independent reproduction.

### ODAM

Instance IoU: 47.78. Result ID: 6a325254fbc373592c2fa655.

[Reporting paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

Recorded evaluation configuration: {"Model":"Grounding DINO"}. This record does not assert independent reproduction.

### SSGrad-CAM++

Instance IoU: 41.13. Result ID: 6a325254fbc373592c2fa652.

[Reporting paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

Recorded evaluation configuration: {"Model":"Grounding DINO"}. This record does not assert independent reproduction.

### Grad-CAM

Instance IoU: 37.49. Result ID: 6a325254fbc373592c2fa651.

[Reporting paper](<https://www.sota2.com/research/paper/6a322a57d33dfb23e783ed5d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66cb1e209131cea12568.md>)

[Implementation](<https://github.com/ramprs/grad-cam/>)

Recorded evaluation configuration: {"Model":"Grounding DINO"}. This record does not assert independent reproduction.

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