# Correlation with human judgments on Flickr8k Expert (tau\_b)

Source: [Correlation with human judgments on Flickr8k Expert (tau\_b)](<https://www.sota2.com/research/sota/correlation-with-human-judgments-on-flickr8k-expert-tau_b>)

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

Data updated: 2026-04-21T09:05:11.667Z

Evaluation of model output correlation with human judgment scores on the Flickr8k Expert dataset, specifically measured using Kendall's Tau-b.

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 | Correlation with human judgments |
| Dataset (including any stated split) | Flickr8k Expert |
| Metric | Tau-b Correlation |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 19 of 19 |
| 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

- [Tau-b Correlation](<https://www.sota2.com/research/sota/correlation-with-human-judgments-on-flickr8k-expert-tau_b.md?metric=Tau-b%20Correlation>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Tau-b Correlation | Result date | Report paper |
| --- | --- | --- | --- |
| VC-Inspector-7B | 63.4 | Not recorded | 68fa7b6e1e209131cea88652 |
| VC-Inspector-3B | 59.9 | Not recorded | 68fa7b6e1e209131cea88652 |
| HICE-S | 57.2 | Not recorded | 68fa7b6e1e209131cea88652 |
| Polos | 56.4 | Not recorded | 68fa7b6e1e209131cea88652 |
| HICE-S | 55.9 | Not recorded | 68fa7b6e1e209131cea88652 |
| PAC-S | 55.5 | Not recorded | 68fa7b6e1e209131cea88652 |
| PAC-S++ | 55.3 | Not recorded | 68fa7b6e1e209131cea88652 |
| PAC-S++ | 54.1 | Not recorded | 68fa7b6e1e209131cea88652 |
| PAC-S | 53.9 | Not recorded | 68fa7b6e1e209131cea88652 |
| FLEUR | 52.7 | Not recorded | 68fa7b6e1e209131cea88652 |
| CLIPScore | 52.6 | Not recorded | 68fa7b6e1e209131cea88652 |
| SPICE | 51.7 | Not recorded | 68fa7b6e1e209131cea88652 |
| FLEUR | 51.6 | Not recorded | 68fa7b6e1e209131cea88652 |
| CLIPScore | 51.1 | Not recorded | 68fa7b6e1e209131cea88652 |
| CIDEr | 43.6 | Not recorded | 68fa7b6e1e209131cea88652 |
| METEOR | 41.5 | Not recorded | 68fa7b6e1e209131cea88652 |
| BLEU\_1 | 32.2 | Not recorded | 68fa7b6e1e209131cea88652 |
| ROUGE | 32.1 | Not recorded | 68fa7b6e1e209131cea88652 |
| BLEU\_4 | 30.6 | Not recorded | 68fa7b6e1e209131cea88652 |

### VC-Inspector-7B

Tau-b Correlation: 63.4. Result ID: 69e73caf47f34ce924cda33e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-free"}. This record does not assert independent reproduction.

### VC-Inspector-3B

Tau-b Correlation: 59.9. Result ID: 69e73caf47f34ce924cda33d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-free"}. This record does not assert independent reproduction.

### HICE-S

Tau-b Correlation: 57.2. Result ID: 69e73caf47f34ce924cda335.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74421e209131cea6096b.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

### Polos

Tau-b Correlation: 56.4. Result ID: 69e73caf47f34ce924cda337.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72551e209131cea55b08.md>)

[Implementation](<https://yuiga.dev/polos>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

### HICE-S

Tau-b Correlation: 55.9. Result ID: 69e73caf47f34ce924cda33b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74421e209131cea6096b.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-free"}. This record does not assert independent reproduction.

### PAC-S

Tau-b Correlation: 55.5. Result ID: 69e73caf47f34ce924cda333.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70201e209131cea493e8.md>)

[Implementation](<https://github.com/aimagelab/pacscore>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

### PAC-S++

Tau-b Correlation: 55.3. Result ID: 69e73caf47f34ce924cda336.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a441e209131cea820a0.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based","Backbone":"ViT-B"}. This record does not assert independent reproduction.

### PAC-S++

Tau-b Correlation: 54.1. Result ID: 69e73caf47f34ce924cda33c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a441e209131cea820a0.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-free","Backbone":"ViT-B"}. This record does not assert independent reproduction.

### PAC-S

Tau-b Correlation: 53.9. Result ID: 69e73caf47f34ce924cda339.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70201e209131cea493e8.md>)

[Implementation](<https://github.com/aimagelab/pacscore>)

Recorded evaluation configuration: {"Reference Setting":"Reference-free"}. This record does not assert independent reproduction.

### FLEUR

Tau-b Correlation: 52.7. Result ID: 69e73caf47f34ce924cda33a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73881e209131cea5d06e.md>)

[Implementation](<https://github.com/Yebin46/FLEUR>)

Recorded evaluation configuration: {"Reference Setting":"Reference-free"}. This record does not assert independent reproduction.

### CLIPScore

Tau-b Correlation: 52.6. Result ID: 69e73caf47f34ce924cda332.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c0f1e209131cea31cc9.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

### SPICE

Tau-b Correlation: 51.7. Result ID: 69e73caf47f34ce924cda331.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62a41e209131ce9fa82e.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

### FLEUR

Tau-b Correlation: 51.6. Result ID: 69e73caf47f34ce924cda334.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73881e209131cea5d06e.md>)

[Implementation](<https://github.com/Yebin46/FLEUR>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

### CLIPScore

Tau-b Correlation: 51.1. Result ID: 69e73caf47f34ce924cda338.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c0f1e209131cea31cc9.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-free"}. This record does not assert independent reproduction.

### CIDEr

Tau-b Correlation: 43.6. Result ID: 69e73caf47f34ce924cda330.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa61e81e209131ce9f6a37.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

### METEOR

Tau-b Correlation: 41.5. Result ID: 69e73caf47f34ce924cda32f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

### BLEU\_1

Tau-b Correlation: 32.2. Result ID: 69e73caf47f34ce924cda32c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

### ROUGE

Tau-b Correlation: 32.1. Result ID: 69e73caf47f34ce924cda32e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

### BLEU\_4

Tau-b Correlation: 30.6. Result ID: 69e73caf47f34ce924cda32d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b6e1e209131cea88652.md>)

Recorded evaluation configuration: {"Reference Setting":"Reference-based"}. This record does not assert independent reproduction.

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