# Thermal-to-Text Retrieval (via Vision pivot) on FLIR to LLVIP (test)

Source: [Thermal-to-Text Retrieval (via Vision pivot) on FLIR to LLVIP (test)](<https://www.sota2.com/research/sota/thermal-to-text-retrieval-via-vision-pivot-on-flir-to-llvip-test>)

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

Data updated: 2026-02-26T06:09:43.366Z

Evaluation of Thermal-to-Text Retrieval performance using a Vision pivot approach on the test split of the FLIR-to-LLVIP cross-dataset benchmark, reporting Mean Average Precision (mAP).

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 | Thermal-to-Text Retrieval (via Vision pivot) |
| Dataset (including any stated split) | FLIR to LLVIP (test) |
| Metric | mAP |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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

- [mAP](<https://www.sota2.com/research/sota/thermal-to-text-retrieval-via-vision-pivot-on-flir-to-llvip-test.md?metric=mAP>): Higher is better; unit: PERCENT.

## Recorded results

| Method | mAP | Result date | Report paper |
| --- | --- | --- | --- |
| LB+BB | 45.2 | Not recorded | 698969e1c97a0b3e3b5dc383 |
| IB+BB | 42.5 | Not recorded | 698969e1c97a0b3e3b5dc383 |
| LB+FT | 23.4 | Not recorded | 698969e1c97a0b3e3b5dc383 |
| IB+FT | 21.2 | Not recorded | 698969e1c97a0b3e3b5dc383 |
| IB | 15.8 | Not recorded | 698969e1c97a0b3e3b5dc383 |
| LB | 14.7 | Not recorded | 698969e1c97a0b3e3b5dc383 |

### LB+BB

mAP: 45.2. Result ID: 6998e85b5698665ec44cc427.

[Reporting paper](<https://www.sota2.com/research/paper/698969e1c97a0b3e3b5dc383.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698969e1c97a0b3e3b5dc383.md>)

Recorded evaluation configuration: {"Backbone":"LanguageBind","Variant":"BrokenBind"}. This record does not assert independent reproduction.

### IB+BB

mAP: 42.5. Result ID: 6998e85b5698665ec44cc424.

[Reporting paper](<https://www.sota2.com/research/paper/698969e1c97a0b3e3b5dc383.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698969e1c97a0b3e3b5dc383.md>)

Recorded evaluation configuration: {"Backbone":"ImageBind","Variant":"BrokenBind"}. This record does not assert independent reproduction.

### LB+FT

mAP: 23.4. Result ID: 6998e85b5698665ec44cc426.

[Reporting paper](<https://www.sota2.com/research/paper/698969e1c97a0b3e3b5dc383.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71f31e209131cea53b97.md>)

[Implementation](<https://github.com/PKU-YuanGroup/LanguageBind>)

Recorded evaluation configuration: {"Backbone":"LanguageBind","Variant":"Fine-tuned"}. This record does not assert independent reproduction.

### IB+FT

mAP: 21.2. Result ID: 6998e85b5698665ec44cc423.

[Reporting paper](<https://www.sota2.com/research/paper/698969e1c97a0b3e3b5dc383.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f971e209131cea467c4.md>)

[Implementation](<https://facebookresearch.github.io/ImageBind>)

Recorded evaluation configuration: {"Backbone":"ImageBind","Variant":"Fine-tuned"}. This record does not assert independent reproduction.

### IB

mAP: 15.8. Result ID: 6998e85b5698665ec44cc422.

[Reporting paper](<https://www.sota2.com/research/paper/698969e1c97a0b3e3b5dc383.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f971e209131cea467c4.md>)

[Implementation](<https://facebookresearch.github.io/ImageBind>)

Recorded evaluation configuration: {"Backbone":"ImageBind","Variant":"Zero-shot"}. This record does not assert independent reproduction.

### LB

mAP: 14.7. Result ID: 6998e85b5698665ec44cc425.

[Reporting paper](<https://www.sota2.com/research/paper/698969e1c97a0b3e3b5dc383.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71f31e209131cea53b97.md>)

[Implementation](<https://github.com/PKU-YuanGroup/LanguageBind>)

Recorded evaluation configuration: {"Backbone":"LanguageBind","Variant":"Zero-shot"}. This record does not assert independent reproduction.

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