# Zero-shot Object Detection on LVIS rare V1 (undetected failure samples)

Source: [Zero-shot Object Detection on LVIS rare V1 (undetected failure samples)](<https://www.sota2.com/research/sota/zero-shot-object-detection-on-lvis-rare-v1-undetected-failure-samples>)

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

Data updated: 2026-06-17T07:52:54.912Z

Evaluation of Zero-shot Object Detection performance specifically on the subset of LVIS v1 rare categories identified as undetected failure samples, utilizing specialized metrics like ESR and MECave.

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 | Zero-shot Object Detection |
| Dataset (including any stated split) | LVIS rare V1 (undetected failure samples) |
| Metric | Instance Metric |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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 Metric](<https://www.sota2.com/research/sota/zero-shot-object-detection-on-lvis-rare-v1-undetected-failure-samples.md?metric=Instance%20Metric>): Higher is better; unit: PERCENT.

- [Instance Metric (Class-level)](<https://www.sota2.com/research/sota/zero-shot-object-detection-on-lvis-rare-v1-undetected-failure-samples.md?metric=Instance%20Metric%20(Class-level)>): Higher is better; unit: PERCENT.

- [Average High Score](<https://www.sota2.com/research/sota/zero-shot-object-detection-on-lvis-rare-v1-undetected-failure-samples.md?metric=Average%20High%20Score>): Higher is better; unit: PERCENT.

- [ESR](<https://www.sota2.com/research/sota/zero-shot-object-detection-on-lvis-rare-v1-undetected-failure-samples.md?metric=ESR>): Lower is better; unit: PERCENT.

- [MECave](<https://www.sota2.com/research/sota/zero-shot-object-detection-on-lvis-rare-v1-undetected-failure-samples.md?metric=MECave>): Higher is better; unit: SCALAR.

- [A-C Ratio](<https://www.sota2.com/research/sota/zero-shot-object-detection-on-lvis-rare-v1-undetected-failure-samples.md?metric=A-C%20Ratio>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Instance Metric | Result date | Report paper |
| --- | --- | --- | --- |
| Greedy | 11.55 | Not recorded | 6a322a57d33dfb23e783ed5d |
| PhaseWin | 9.42 | Not recorded | 6a322a57d33dfb23e783ed5d |
| PhaseWin | 7.87 | Not recorded | 6a322a57d33dfb23e783ed5d |
| RISE | 7.03 | Not recorded | 6a322a57d33dfb23e783ed5d |
| HSIC | 5.16 | Not recorded | 6a322a57d33dfb23e783ed5d |
| SSGrad-CAM++ | 2.92 | Not recorded | 6a322a57d33dfb23e783ed5d |
| Grad-CAM | 2.91 | Not recorded | 6a322a57d33dfb23e783ed5d |
| ODAM | 2.83 | Not recorded | 6a322a57d33dfb23e783ed5d |

### Greedy

Instance Metric: 11.55. Result ID: 6a325254fbc373592c2fa61f.

[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: {"Backbone":"Grounding DINO","Steps":"100"}. This record does not assert independent reproduction.

### PhaseWin

Instance Metric: 9.42. Result ID: 6a325254fbc373592c2fa621.

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

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

Recorded evaluation configuration: {"Backbone":"Grounding DINO","Steps":"100"}. This record does not assert independent reproduction.

### PhaseWin

Instance Metric: 7.87. Result ID: 6a325254fbc373592c2fa620.

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

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

Recorded evaluation configuration: {"Backbone":"Grounding DINO","Steps":"50"}. This record does not assert independent reproduction.

### RISE

Instance Metric: 7.03. Result ID: 6a325254fbc373592c2fa61c.

[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: {"Backbone":"Grounding DINO"}. This record does not assert independent reproduction.

### HSIC

Instance Metric: 5.16. Result ID: 6a325254fbc373592c2fa61d.

[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: {"Backbone":"Grounding DINO"}. This record does not assert independent reproduction.

### SSGrad-CAM++

Instance Metric: 2.92. Result ID: 6a325254fbc373592c2fa61b.

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

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

### Grad-CAM

Instance Metric: 2.91. Result ID: 6a325254fbc373592c2fa61a.

[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: {"Backbone":"Grounding DINO"}. This record does not assert independent reproduction.

### ODAM

Instance Metric: 2.83. Result ID: 6a325254fbc373592c2fa61e.

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

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

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