# Generalized Zero-Shot Classification on AudioSet ZSL (test)

Source: [Generalized Zero-Shot Classification on AudioSet ZSL (test)](<https://www.sota2.com/research/sota/generalized-zero-shot-classification-on-audioset-zsl-test>)

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

Data updated: 2026-02-26T06:10:06.200Z

Evaluation of Generalized Zero-Shot Classification performance on the test split of the AudioSet ZSL dataset, measured using mAcc scores for Seen (S), Unseen (U), and Harmonic Mean (HM).

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 | Generalized Zero-Shot Classification |
| Dataset (including any stated split) | AudioSet ZSL (test) |
| Metric | mAcc (Seen) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 16 of 16 |
| 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

- [mAcc (Seen)](<https://www.sota2.com/research/sota/generalized-zero-shot-classification-on-audioset-zsl-test.md?metric=mAcc%20(Seen)>): Higher is better; unit: PERCENT.

- [mAcc (Unseen)](<https://www.sota2.com/research/sota/generalized-zero-shot-classification-on-audioset-zsl-test.md?metric=mAcc%20(Unseen)>): Higher is better; unit: PERCENT.

- [mAcc (HM)](<https://www.sota2.com/research/sota/generalized-zero-shot-classification-on-audioset-zsl-test.md?metric=mAcc%20(HM)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | mAcc (Seen) | Result date | Report paper |
| --- | --- | --- | --- |
| AVGZSLNet | 50.96 | Not recorded | 68fa68c11e209131cea1e112 |
| CONSE | 48.5 | Not recorded | 68fa68c11e209131cea1e112 |
| AVGZSLNet | 48.05 | Not recorded | 68fa68c11e209131cea1e112 |
| ALE | 47.9 | Not recorded | 68fa68c11e209131cea1e112 |
| AVGZSLNet | 46.04 | Not recorded | 68fa68c11e209131cea1e112 |
| audio-video concat model | 45.83 | Not recorded | 68fa68c11e209131cea1e112 |
| video only model | 43.27 | Not recorded | 68fa68c11e209131cea1e112 |
| CJME | 41.53 | Not recorded | 68fa68c11e209131cea1e112 |
| CJME | 41.07 | Not recorded | 68fa68c11e209131cea1e112 |
| DEVISE | 39.8 | Not recorded | 68fa68c11e209131cea1e112 |
| ESZSL | 33.8 | Not recorded | 68fa68c11e209131cea1e112 |
| AVGZSLNet | 33.29 | Not recorded | 68fa68c11e209131cea1e112 |
| CJME | 30.29 | Not recorded | 68fa68c11e209131cea1e112 |
| SAE | 29.3 | Not recorded | 68fa68c11e209131cea1e112 |
| audio only model | 28.35 | Not recorded | 68fa68c11e209131cea1e112 |
| CJME | 25.58 | Not recorded | 68fa68c11e209131cea1e112 |

### AVGZSLNet

mAcc (Seen): 50.96. Result ID: 698d82e18b517eb433220cf0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Implementation](<https://github.com/krantiparida/AudioSetZSL>)

Recorded evaluation configuration: {"Test Modality":"both","fusion":"with attention"}. This record does not assert independent reproduction.

### CONSE

mAcc (Seen): 48.5. Result ID: 698d82e18b517eb433220ce5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa632f1e209131ce9fd1fb.md>)

Recorded evaluation configuration: {"Test Modality":"video"}. This record does not assert independent reproduction.

### AVGZSLNet

mAcc (Seen): 48.05. Result ID: 698d82e18b517eb433220ceb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Implementation](<https://github.com/krantiparida/AudioSetZSL>)

Recorded evaluation configuration: {"Test Modality":"video"}. This record does not assert independent reproduction.

### ALE

mAcc (Seen): 47.9. Result ID: 698d82e18b517eb433220ce9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62d31e209131ce9fb5fb.md>)

[Implementation](<https://code.google.com/p/word2vec/>)

Recorded evaluation configuration: {"Test Modality":"video"}. This record does not assert independent reproduction.

### AVGZSLNet

mAcc (Seen): 46.04. Result ID: 698d82e18b517eb433220cee.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Implementation](<https://github.com/krantiparida/AudioSetZSL>)

Recorded evaluation configuration: {"Test Modality":"both","fusion":"equal weight"}. This record does not assert independent reproduction.

### audio-video concat model

mAcc (Seen): 45.83. Result ID: 698d82e18b517eb433220cec.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

Recorded evaluation configuration: {"Test Modality":"both"}. This record does not assert independent reproduction.

### video only model

mAcc (Seen): 43.27. Result ID: 698d82e18b517eb433220ce4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66931e209131cea1106d.md>)

[Implementation](<https://www.cse.iitk.ac.in/users/kranti/avzsl.html>)

Recorded evaluation configuration: {"Test Modality":"video"}. This record does not assert independent reproduction.

### CJME

mAcc (Seen): 41.53. Result ID: 698d82e18b517eb433220cea.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66931e209131cea1106d.md>)

[Implementation](<https://www.cse.iitk.ac.in/users/kranti/avzsl.html>)

Recorded evaluation configuration: {"Test Modality":"video"}. This record does not assert independent reproduction.

### CJME

mAcc (Seen): 41.07. Result ID: 698d82e18b517eb433220cef.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66931e209131cea1106d.md>)

[Implementation](<https://www.cse.iitk.ac.in/users/kranti/avzsl.html>)

Recorded evaluation configuration: {"Test Modality":"both","fusion":"with attention"}. This record does not assert independent reproduction.

### DEVISE

mAcc (Seen): 39.8. Result ID: 698d82e18b517eb433220ce6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

Recorded evaluation configuration: {"Test Modality":"video"}. This record does not assert independent reproduction.

### ESZSL

mAcc (Seen): 33.8. Result ID: 698d82e18b517eb433220ce8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

Recorded evaluation configuration: {"Test Modality":"video"}. This record does not assert independent reproduction.

### AVGZSLNet

mAcc (Seen): 33.29. Result ID: 698d82e18b517eb433220ce3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Implementation](<https://github.com/krantiparida/AudioSetZSL>)

Recorded evaluation configuration: {"Test Modality":"audio"}. This record does not assert independent reproduction.

### CJME

mAcc (Seen): 30.29. Result ID: 698d82e18b517eb433220ced.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66931e209131cea1106d.md>)

[Implementation](<https://www.cse.iitk.ac.in/users/kranti/avzsl.html>)

Recorded evaluation configuration: {"Test Modality":"both","fusion":"equal weight"}. This record does not assert independent reproduction.

### SAE

mAcc (Seen): 29.3. Result ID: 698d82e18b517eb433220ce7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa634b1e209131ce9fdc97.md>)

Recorded evaluation configuration: {"Test Modality":"video"}. This record does not assert independent reproduction.

### audio only model

mAcc (Seen): 28.35. Result ID: 698d82e18b517eb433220ce1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66931e209131cea1106d.md>)

[Implementation](<https://www.cse.iitk.ac.in/users/kranti/avzsl.html>)

Recorded evaluation configuration: {"Test Modality":"audio"}. This record does not assert independent reproduction.

### CJME

mAcc (Seen): 25.58. Result ID: 698d82e18b517eb433220ce2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68c11e209131cea1e112.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66931e209131cea1106d.md>)

[Implementation](<https://www.cse.iitk.ac.in/users/kranti/avzsl.html>)

Recorded evaluation configuration: {"Test Modality":"audio"}. This record does not assert independent reproduction.

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