# Speech Emotion Recognition on IEMOCAP Speaker-dependent random train-test split

Source: [Speech Emotion Recognition on IEMOCAP Speaker-dependent random train-test split](<https://www.sota2.com/research/sota/speech-emotion-recognition-on-iemocap-speaker-dependent-random-train-test-split>)

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

Data updated: 2026-02-26T06:08:21.534Z

Evaluation of Speech Emotion Recognition performance on the speaker-dependent random train-test split of the IEMOCAP dataset, reporting Weighted Accuracy (WA).

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 | Speech Emotion Recognition |
| Dataset (including any stated split) | IEMOCAP Speaker-dependent random train-test split |
| Metric | WA |
| 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

- [WA](<https://www.sota2.com/research/sota/speech-emotion-recognition-on-iemocap-speaker-dependent-random-train-test-split.md?metric=WA>): Higher is better; unit: PERCENT.

## Recorded results

| Method | WA | Result date | Report paper |
| --- | --- | --- | --- |
| HuBERT Large + Ours (TAP) | 81 | Not recorded | 68fa6dd31e209131cea3c35b |
| HuBERT Large + Ours (SNP) | 79.9 | Not recorded | 68fa6dd31e209131cea3c35b |
| HuBERT Large | 78.3 | Not recorded | 68fa6dd31e209131cea3c35b |
| MOMA | 74.8 | Not recorded | 68fa6dd31e209131cea3c35b |
| HuBERT Base + Ours (SNP) | 72.8 | Not recorded | 68fa6dd31e209131cea3c35b |
| DS-LSTM | 72.7 | Not recorded | 68fa6dd31e209131cea3c35b |
| HuBERT Base + Ours (TAP) | 72.4 | Not recorded | 68fa6dd31e209131cea3c35b |
| MDRE | 71.8 | Not recorded | 68fa6dd31e209131cea3c35b |
| HuBERT Base | 71.2 | Not recorded | 68fa6dd31e209131cea3c35b |

### HuBERT Large + Ours (TAP)

WA: 81. Result ID: 698fea1b3a6c314849fcc17d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

Recorded evaluation configuration: {"Backbone":"HuBERT Large","Strategy":"Train all parameters (TAP)"}. This record does not assert independent reproduction.

### HuBERT Large + Ours (SNP)

WA: 79.9. Result ID: 698fea1b3a6c314849fcc17c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

Recorded evaluation configuration: {"Backbone":"HuBERT Large","Strategy":"Speaker normalization projector (SNP)"}. This record does not assert independent reproduction.

### HuBERT Large

WA: 78.3. Result ID: 698fea1b3a6c314849fcc17b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a2a1e209131cea26880.md>)

[Implementation](<https://github.com/pytorch/fairseq/tree/master/examples/hubert>)

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

### MOMA

WA: 74.8. Result ID: 698fea1b3a6c314849fcc177.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### HuBERT Base + Ours (SNP)

WA: 72.8. Result ID: 698fea1b3a6c314849fcc179.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

Recorded evaluation configuration: {"Backbone":"HuBERT Base","Strategy":"Speaker normalization projector (SNP)"}. This record does not assert independent reproduction.

### DS-LSTM

WA: 72.7. Result ID: 698fea1b3a6c314849fcc176.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa680c1e209131cea19d2f.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### HuBERT Base + Ours (TAP)

WA: 72.4. Result ID: 698fea1b3a6c314849fcc17a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

Recorded evaluation configuration: {"Backbone":"HuBERT Base","Strategy":"Train all parameters (TAP)"}. This record does not assert independent reproduction.

### MDRE

WA: 71.8. Result ID: 698fea1b3a6c314849fcc175.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64e21e209131cea0717e.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### HuBERT Base

WA: 71.2. Result ID: 698fea1b3a6c314849fcc178.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6dd31e209131cea3c35b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a2a1e209131cea26880.md>)

[Implementation](<https://github.com/pytorch/fairseq/tree/master/examples/hubert>)

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

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