# Automatic Speech Recognition on CHiME-4 single-channel (et\_simu)

Source: [Automatic Speech Recognition on CHiME-4 single-channel (et\_simu)](<https://www.sota2.com/research/sota/automatic-speech-recognition-on-chime-4-single-channel-et_simu>)

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

Data updated: 2026-02-26T06:08:31.790Z

Evaluation of Automatic Speech Recognition performance on the single-channel simulated evaluation split (et\_simu) of the CHiME-4 dataset, typically measured using Word Error Rate (WER).

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 | Automatic Speech Recognition |
| Dataset (including any stated split) | CHiME-4 single-channel (et\_simu) |
| Metric | WER |
| Unit | SCALAR |
| Direction | Lower 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

- [WER](<https://www.sota2.com/research/sota/automatic-speech-recognition-on-chime-4-single-channel-et_simu.md?metric=WER>): Lower is better; unit: SCALAR.

## Recorded results

| Method | WER | Result date | Report paper |
| --- | --- | --- | --- |
| WavLM (Weighted-sum) + Joint FT. WavLM | 0.056 | Not recorded | 6984246ba8b01e76cc8ee973 |
| IRIS | 0.061 | Not recorded | 6984246ba8b01e76cc8ee973 |
| WavLM (Token) + W2T-E | 0.061 | Not recorded | 6984246ba8b01e76cc8ee973 |
| WavLM (Weighted-sum) | 0.08 | Not recorded | 6984246ba8b01e76cc8ee973 |
| WavLM (Weighted-sum) + W2W-E (TF-GridNet) | 0.082 | Not recorded | 6984246ba8b01e76cc8ee973 |
| W2T-E | 0.082 | Not recorded | 6984246ba8b01e76cc8ee973 |
| W2T-E | 0.099 | Not recorded | 6984246ba8b01e76cc8ee973 |
| WavLM (Weighted-sum) | 0.11 | Not recorded | 6984246ba8b01e76cc8ee973 |
| W2W-E (TF-GridNet) | 0.112 | Not recorded | 6984246ba8b01e76cc8ee973 |
| V2T-E (E-Branchformer) | 0.136 | Not recorded | 6984246ba8b01e76cc8ee973 |
| V2T-E (TCN) | 0.137 | Not recorded | 6984246ba8b01e76cc8ee973 |
| V2T-E (MLP) | 0.147 | Not recorded | 6984246ba8b01e76cc8ee973 |
| W2W-E (TF-GridNet) | 0.151 | Not recorded | 6984246ba8b01e76cc8ee973 |
| W2W-E (Conv-TasNet) | 0.172 | Not recorded | 6984246ba8b01e76cc8ee973 |
| WavLM | 0.186 | Not recorded | 6984246ba8b01e76cc8ee973 |
| T2T-E | 0.186 | Not recorded | 6984246ba8b01e76cc8ee973 |
| W2W-E (Conv-TasNet) | 0.216 | Not recorded | 6984246ba8b01e76cc8ee973 |
| WavLM | 0.236 | Not recorded | 6984246ba8b01e76cc8ee973 |
| FBANK | 0.251 | Not recorded | 6984246ba8b01e76cc8ee973 |

### WavLM (Weighted-sum) + Joint FT. WavLM

WER: 0.056. Result ID: 6999fe4a5698665ec4801f29.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dfa1e209131cea3d210.md>)

[Implementation](<https://aka.ms/wavlm>)

Recorded evaluation configuration: {"Decoding":"LM-based","Training":"Joint Fine-tuning"}. This record does not assert independent reproduction.

### IRIS

WER: 0.061. Result ID: 6999fe4a5698665ec4801f26.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c311e209131cea32bf7.md>)

Recorded evaluation configuration: {"Decoding":"LM-based","Architecture":"Conv-TasNet + AED","Backbone":"WavLM Large"}. This record does not assert independent reproduction.

### WavLM (Token) + W2T-E

WER: 0.061. Result ID: 6999fe4a5698665ec4801f2a.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

Recorded evaluation configuration: {"Decoding":"LM-based","Frontend":"W2T-E"}. This record does not assert independent reproduction.

### WavLM (Weighted-sum)

WER: 0.08. Result ID: 6999fe4a5698665ec4801f27.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dfa1e209131cea3d210.md>)

[Implementation](<https://aka.ms/wavlm>)

Recorded evaluation configuration: {"Decoding":"LM-based","Aggregation":"Weighted-sum"}. This record does not assert independent reproduction.

### WavLM (Weighted-sum) + W2W-E (TF-GridNet)

WER: 0.082. Result ID: 6999fe4a5698665ec4801f28.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70361e209131cea49961.md>)

Recorded evaluation configuration: {"Decoding":"LM-based","Frontend":"W2W-E (TF-GridNet)"}. This record does not assert independent reproduction.

### W2T-E

WER: 0.082. Result ID: 6999fe4a5698665ec4801f5b.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

Recorded evaluation configuration: {"System":"+ W2T-E","Backend Architecture":"Token ASR (AED)"}. This record does not assert independent reproduction.

### W2T-E

WER: 0.099. Result ID: 6999fe4a5698665ec4801f5d.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

Recorded evaluation configuration: {"System":"+ W2T-E","Backend Architecture":"Token ASR (CTC-only)"}. This record does not assert independent reproduction.

### WavLM (Weighted-sum)

WER: 0.11. Result ID: 6999fe4a5698665ec4801f51.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dfa1e209131cea3d210.md>)

[Implementation](<https://aka.ms/wavlm>)

Recorded evaluation configuration: {"System":"WavLM (Weighted-sum)","Backend Architecture":"Continuous ASR (AED)"}. This record does not assert independent reproduction.

### W2W-E (TF-GridNet)

WER: 0.112. Result ID: 6999fe4a5698665ec4801f53.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70361e209131cea49961.md>)

Recorded evaluation configuration: {"System":"+ W2W-E (TF-GridNet)","Backend Architecture":"Continuous ASR (AED)"}. This record does not assert independent reproduction.

### V2T-E (E-Branchformer)

WER: 0.136. Result ID: 6999fe4a5698665ec4801f5a.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

Recorded evaluation configuration: {"System":"+ V2T-E (E-Branchformer)","Backend Architecture":"Token ASR (AED)"}. This record does not assert independent reproduction.

### V2T-E (TCN)

WER: 0.137. Result ID: 6999fe4a5698665ec4801f59.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

Recorded evaluation configuration: {"System":"+ V2T-E (TCN)","Backend Architecture":"Token ASR (AED)"}. This record does not assert independent reproduction.

### V2T-E (MLP)

WER: 0.147. Result ID: 6999fe4a5698665ec4801f58.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

Recorded evaluation configuration: {"System":"+ V2T-E (MLP)","Backend Architecture":"Token ASR (AED)"}. This record does not assert independent reproduction.

### W2W-E (TF-GridNet)

WER: 0.151. Result ID: 6999fe4a5698665ec4801f56.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70361e209131cea49961.md>)

Recorded evaluation configuration: {"System":"+ W2W-E (TF-GridNet)","Backend Architecture":"Token ASR (AED)"}. This record does not assert independent reproduction.

### W2W-E (Conv-TasNet)

WER: 0.172. Result ID: 6999fe4a5698665ec4801f52.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65bc1e209131cea0c05a.md>)

[Implementation](<https://github.com/naplab/Conv-TasNet>)

Recorded evaluation configuration: {"System":"+ W2W-E (Conv-TasNet)","Backend Architecture":"Continuous ASR (AED)"}. This record does not assert independent reproduction.

### WavLM

WER: 0.186. Result ID: 6999fe4a5698665ec4801f54.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dfa1e209131cea3d210.md>)

[Implementation](<https://aka.ms/wavlm>)

Recorded evaluation configuration: {"System":"WavLM","Backend Architecture":"Token ASR (AED)"}. This record does not assert independent reproduction.

### T2T-E

WER: 0.186. Result ID: 6999fe4a5698665ec4801f57.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

Recorded evaluation configuration: {"System":"+ T2T-E","Backend Architecture":"Token ASR (AED)"}. This record does not assert independent reproduction.

### W2W-E (Conv-TasNet)

WER: 0.216. Result ID: 6999fe4a5698665ec4801f55.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65bc1e209131cea0c05a.md>)

[Implementation](<https://github.com/naplab/Conv-TasNet>)

Recorded evaluation configuration: {"System":"+ W2W-E (Conv-TasNet)","Backend Architecture":"Token ASR (AED)"}. This record does not assert independent reproduction.

### WavLM

WER: 0.236. Result ID: 6999fe4a5698665ec4801f5c.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dfa1e209131cea3d210.md>)

[Implementation](<https://aka.ms/wavlm>)

Recorded evaluation configuration: {"System":"WavLM","Backend Architecture":"Token ASR (CTC-only)"}. This record does not assert independent reproduction.

### FBANK

WER: 0.251. Result ID: 6999fe4a5698665ec4801f50.

[Reporting paper](<https://www.sota2.com/research/paper/6984246ba8b01e76cc8ee973.md>)

Recorded evaluation configuration: {"System":"FBANK","Backend Architecture":"Continuous ASR (AED)"}. This record does not assert independent reproduction.

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