# Open-Vocabulary Object Detection on FG-OVD (Difficulty and Attribute Metrics)

Source: [Open-Vocabulary Object Detection on FG-OVD (Difficulty and Attribute Metrics)](<https://www.sota2.com/research/sota/open-vocabulary-object-detection-on-fg-ovd-difficulty-and-attribute-metrics>)

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

Data updated: 2026-05-19T12:04:06.481Z

Evaluation of Open-Vocabulary Object Detection performance on the FG-OVD dataset, reporting metrics broken down by difficulty levels (Easy, Medium, Hard, Trivial) and specific visual attributes (Color, Material, Pattern, Transparency).

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 | Open-Vocabulary Object Detection |
| Dataset (including any stated split) | FG-OVD |
| Metric | Score (Hard Difficulty) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 16 of 16 |
| 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

- [Score (Hard Difficulty)](<https://www.sota2.com/research/sota/open-vocabulary-object-detection-on-fg-ovd-difficulty-and-attribute-metrics.md?metric=Score%20(Hard%20Difficulty)>): Higher is better; unit: PERCENT.

- [Score (Medium Difficulty)](<https://www.sota2.com/research/sota/open-vocabulary-object-detection-on-fg-ovd-difficulty-and-attribute-metrics.md?metric=Score%20(Medium%20Difficulty)>): Higher is better; unit: PERCENT.

- [Score (Easy Difficulty)](<https://www.sota2.com/research/sota/open-vocabulary-object-detection-on-fg-ovd-difficulty-and-attribute-metrics.md?metric=Score%20(Easy%20Difficulty)>): Higher is better; unit: PERCENT.

- [Score (Trivial Difficulty)](<https://www.sota2.com/research/sota/open-vocabulary-object-detection-on-fg-ovd-difficulty-and-attribute-metrics.md?metric=Score%20(Trivial%20Difficulty)>): Higher is better; unit: PERCENT.

- [Score (Color Attribute)](<https://www.sota2.com/research/sota/open-vocabulary-object-detection-on-fg-ovd-difficulty-and-attribute-metrics.md?metric=Score%20(Color%20Attribute)>): Higher is better; unit: PERCENT.

- [Score (Material Attribute)](<https://www.sota2.com/research/sota/open-vocabulary-object-detection-on-fg-ovd-difficulty-and-attribute-metrics.md?metric=Score%20(Material%20Attribute)>): Higher is better; unit: PERCENT.

- [Score (Pattern Attribute)](<https://www.sota2.com/research/sota/open-vocabulary-object-detection-on-fg-ovd-difficulty-and-attribute-metrics.md?metric=Score%20(Pattern%20Attribute)>): Higher is better; unit: PERCENT.

- [Score (Transparency Attribute)](<https://www.sota2.com/research/sota/open-vocabulary-object-detection-on-fg-ovd-difficulty-and-attribute-metrics.md?metric=Score%20(Transparency%20Attribute)>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/open-vocabulary-object-detection-on-fg-ovd-difficulty-and-attribute-metrics.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Score (Hard Difficulty) | Result date | Report paper |
| --- | --- | --- | --- |
| DesCo + DSAA | 52.9 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| GLIP + DSAA | 47.1 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| DesCo + HA-FGOVD | 45.4 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| DesCo | 43.7 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| GLIP + HA-FGOVD | 40.2 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| Grounding DINO + DSAA | 38.5 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| GLIP | 37.6 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| OWL-ViT(L/14) | 26.5 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| OWL-ViT(B/16) | 26.2 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| OWLv2(L/14) | 25.4 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| OWLv2(B/16) | 25.3 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| ViLD | 22.1 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| Grounding DINO + HA-FGOVD | 19.2 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| Grounding DINO | 17 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| CORA | 13.8 | Not recorded | 6a0beecfdbc836b87d82c6f8 |
| Detic | 11.5 | Not recorded | 6a0beecfdbc836b87d82c6f8 |

### DesCo + DSAA

Score (Hard Difficulty): 52.9. Result ID: 6a0c51b0fbc373592c039d77.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

Recorded evaluation configuration: {"Detector":"DesCo","Enhancement Framework":"DSAA"}. This record does not assert independent reproduction.

### GLIP + DSAA

Score (Hard Difficulty): 47.1. Result ID: 6a0c51b0fbc373592c039d74.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

Recorded evaluation configuration: {"Detector":"GLIP","Enhancement Framework":"DSAA"}. This record does not assert independent reproduction.

### DesCo + HA-FGOVD

Score (Hard Difficulty): 45.4. Result ID: 6a0c51b0fbc373592c039d76.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75531e209131cea672ff.md>)

Recorded evaluation configuration: {"Detector":"DesCo","Enhancement Framework":"HA-FGOVD"}. This record does not assert independent reproduction.

### DesCo

Score (Hard Difficulty): 43.7. Result ID: 6a0c51b0fbc373592c039d75.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6fdb1e209131cea47af8.md>)

Recorded evaluation configuration: {"Detector":"DesCo"}. This record does not assert independent reproduction.

### GLIP + HA-FGOVD

Score (Hard Difficulty): 40.2. Result ID: 6a0c51b0fbc373592c039d73.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75531e209131cea672ff.md>)

Recorded evaluation configuration: {"Detector":"GLIP","Enhancement Framework":"HA-FGOVD"}. This record does not assert independent reproduction.

### Grounding DINO + DSAA

Score (Hard Difficulty): 38.5. Result ID: 6a0c51b0fbc373592c039d7a.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

Recorded evaluation configuration: {"Detector":"Grounding DINO","Enhancement Framework":"DSAA"}. This record does not assert independent reproduction.

### GLIP

Score (Hard Difficulty): 37.6. Result ID: 6a0c51b0fbc373592c039d72.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cd91e209131cea36418.md>)

[Implementation](<https://github.com/microsoft/GLIP>)

Recorded evaluation configuration: {"Detector":"GLIP"}. This record does not assert independent reproduction.

### OWL-ViT(L/14)

Score (Hard Difficulty): 26.5. Result ID: 6a0c51b0fbc373592c039d6c.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

Recorded evaluation configuration: {"Detector":"OWL-ViT(L/14)"}. This record does not assert independent reproduction.

### OWL-ViT(B/16)

Score (Hard Difficulty): 26.2. Result ID: 6a0c51b0fbc373592c039d6b.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

Recorded evaluation configuration: {"Detector":"OWL-ViT(B/16)"}. This record does not assert independent reproduction.

### OWLv2(L/14)

Score (Hard Difficulty): 25.4. Result ID: 6a0c51b0fbc373592c039d6e.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

Recorded evaluation configuration: {"Detector":"OWLv2(L/14)"}. This record does not assert independent reproduction.

### OWLv2(B/16)

Score (Hard Difficulty): 25.3. Result ID: 6a0c51b0fbc373592c039d6d.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

Recorded evaluation configuration: {"Detector":"OWLv2(B/16)"}. This record does not assert independent reproduction.

### ViLD

Score (Hard Difficulty): 22.1. Result ID: 6a0c51b0fbc373592c039d70.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c891e209131cea34901.md>)

[Implementation](<https://github.com/tensorflow/tpu/tree/master/models/official/detection/projects/vild>)

Recorded evaluation configuration: {"Detector":"ViLD"}. This record does not assert independent reproduction.

### Grounding DINO + HA-FGOVD

Score (Hard Difficulty): 19.2. Result ID: 6a0c51b0fbc373592c039d79.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75531e209131cea672ff.md>)

Recorded evaluation configuration: {"Detector":"Grounding DINO","Enhancement Framework":"HA-FGOVD"}. This record does not assert independent reproduction.

### Grounding DINO

Score (Hard Difficulty): 17. Result ID: 6a0c51b0fbc373592c039d78.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74311e209131cea6055d.md>)

[Implementation](<https://github.com/IDEA-Research/GroundingDINO>)

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

### CORA

Score (Hard Difficulty): 13.8. Result ID: 6a0c51b0fbc373592c039d71.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

Recorded evaluation configuration: {"Detector":"CORA"}. This record does not assert independent reproduction.

### Detic

Score (Hard Difficulty): 11.5. Result ID: 6a0c51b0fbc373592c039d6f.

[Reporting paper](<https://www.sota2.com/research/paper/6a0beecfdbc836b87d82c6f8.md>)

Recorded evaluation configuration: {"Detector":"Detic"}. This record does not assert independent reproduction.

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