# Talking Head Generation on RealWorld-LipSync

Source: [Talking Head Generation on RealWorld-LipSync](<https://www.sota2.com/research/sota/talking-head-generation-on-realworld-lipsync>)

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

Data updated: 2026-03-05T08:57:57.011Z

Evaluation of Talking Head Generation models on the RealWorld-LipSync dataset, assessed using metrics covering visual quality (FID, FVD), perceptual quality (HyperIQA, NIQE, BRISQUE), and lip synchronization (CSIM, LSE-C).

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 | Talking Head Generation |
| Dataset (including any stated split) | RealWorld-LipSync |
| Metric | FID |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 7 of 7 |
| 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

- [FID](<https://www.sota2.com/research/sota/talking-head-generation-on-realworld-lipsync.md?metric=FID>): Lower is better; unit: SCALAR.

- [FVD](<https://www.sota2.com/research/sota/talking-head-generation-on-realworld-lipsync.md?metric=FVD>): Lower is better; unit: SCALAR.

- [CSIM](<https://www.sota2.com/research/sota/talking-head-generation-on-realworld-lipsync.md?metric=CSIM>): Higher is better; unit: PERCENT.

- [NIQE](<https://www.sota2.com/research/sota/talking-head-generation-on-realworld-lipsync.md?metric=NIQE>): Lower is better; unit: SCALAR.

- [BRISQUE](<https://www.sota2.com/research/sota/talking-head-generation-on-realworld-lipsync.md?metric=BRISQUE>): Lower is better; unit: SCALAR.

- [HyperIQA](<https://www.sota2.com/research/sota/talking-head-generation-on-realworld-lipsync.md?metric=HyperIQA>): Higher is better; unit: SCALAR.

- [LSE-C](<https://www.sota2.com/research/sota/talking-head-generation-on-realworld-lipsync.md?metric=LSE-C>): Higher is better; unit: SCALAR.

- [GSR](<https://www.sota2.com/research/sota/talking-head-generation-on-realworld-lipsync.md?metric=GSR>): Higher is better; unit: PERCENT.

## Recorded results

| Method | FID | Result date | Report paper |
| --- | --- | --- | --- |
| UniSync | 12.976 | Not recorded | 69a90e132d603e6de3e40889 |
| OmniSync | 13.655 | Not recorded | 69a90e132d603e6de3e40889 |
| LatentSync | 14.039 | Not recorded | 69a90e132d603e6de3e40889 |
| TalkLip | 15.613 | Not recorded | 69a90e132d603e6de3e40889 |
| MuseTalk | 16.894 | Not recorded | 69a90e132d603e6de3e40889 |
| Diff2Lip | 17.631 | Not recorded | 69a90e132d603e6de3e40889 |
| IP-LAP | 19.458 | Not recorded | 69a90e132d603e6de3e40889 |

### UniSync

FID: 12.976. Result ID: 69a945486dbb7075b952dd4f.

[Reporting paper](<https://www.sota2.com/research/paper/69a90e132d603e6de3e40889.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a90e132d603e6de3e40889.md>)

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

### OmniSync

FID: 13.655. Result ID: 69a945486dbb7075b952dd4e.

[Reporting paper](<https://www.sota2.com/research/paper/69a90e132d603e6de3e40889.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b5f1e209131cea87fd2.md>)

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

### LatentSync

FID: 14.039. Result ID: 69a945486dbb7075b952dd4d.

[Reporting paper](<https://www.sota2.com/research/paper/69a90e132d603e6de3e40889.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa776a1e209131cea7323a.md>)

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

### TalkLip

FID: 15.613. Result ID: 69a945486dbb7075b952dd49.

[Reporting paper](<https://www.sota2.com/research/paper/69a90e132d603e6de3e40889.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ef21e209131cea43157.md>)

[Implementation](<https://github.com/Sxjdwang/TalkLip>)

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

### MuseTalk

FID: 16.894. Result ID: 69a945486dbb7075b952dd4c.

[Reporting paper](<https://www.sota2.com/research/paper/69a90e132d603e6de3e40889.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77a01e209131cea74751.md>)

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

### Diff2Lip

FID: 17.631. Result ID: 69a945486dbb7075b952dd4b.

[Reporting paper](<https://www.sota2.com/research/paper/69a90e132d603e6de3e40889.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70571e209131cea4a643.md>)

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

### IP-LAP

FID: 19.458. Result ID: 69a945486dbb7075b952dd4a.

[Reporting paper](<https://www.sota2.com/research/paper/69a90e132d603e6de3e40889.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f561e209131cea451ed.md>)

[Implementation](<https://github.com/Weizhi-Zhong/IP_LAP>)

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

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