# Joint Audio-Video Generation on Custom human-centric audio-video real and anime images (test) Source: [Joint Audio-Video Generation on Custom human-centric audio-video real and anime images (test)]() Collection index: [llms.txt]() Data updated: 2026-03-25T13:36:33.680Z Evaluation of models performing Joint Audio-Video Generation using the test split of the Custom human-centric audio-video real and anime images dataset, assessed using metrics including PQ, CU, and 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 | Joint Audio-Video Generation | | Dataset (including any stated split) | Custom human-centric audio-video real and anime images (test) | | Metric | PQ | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 6 of 6 | | 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 - [PQ](): Higher is better; unit: SCALAR. - [CU](): Higher is better; unit: SCALAR. - [WER](): Lower is better; unit: SCALAR. - [SC](): Higher is better; unit: SCALAR. - [DD](): Lower is better; unit: SCALAR. - [IQ](): Higher is better; unit: SCALAR. - [LS](): Higher is better; unit: SCALAR. - [TC](): Higher is better; unit: SCALAR. - [EC](): Higher is better; unit: SCALAR. ## Recorded results | Method | PQ | Result date | Report paper | | --- | --- | --- | --- | | OmniAvatar | 8.15 | Not recorded | 690c2b732828b7cc6937c495 | | Wan-S2V | 8.15 | Not recorded | 690c2b732828b7cc6937c495 | | UniAVGen | 7 | Not recorded | 690c2b732828b7cc6937c495 | | Ovi | 6.03 | Not recorded | 690c2b732828b7cc6937c495 | | JavisDiT | 5.21 | Not recorded | 690c2b732828b7cc6937c495 | | Universe-1 | 4.56 | Not recorded | 690c2b732828b7cc6937c495 | ### OmniAvatar PQ: 8.15. Result ID: 69c3e40347f34ce924ad6faf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Generation Paradigm":"Two-stage Generation","Parameters (S+V)":"21.1B"}. This record does not assert independent reproduction. ### Wan-S2V PQ: 8.15. Result ID: 69c3e40347f34ce924ad6fb0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Generation Paradigm":"Two-stage Generation","Parameters (S+V)":"16.6B"}. This record does not assert independent reproduction. ### UniAVGen PQ: 7. Result ID: 69c3e40347f34ce924ad6fb4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Generation Paradigm":"Joint Generation","Joint Training Samples":"1.3M","Parameters (S+V)":"7.1B"}. This record does not assert independent reproduction. ### Ovi PQ: 6.03. Result ID: 69c3e40347f34ce924ad6fb3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Generation Paradigm":"Joint Generation","Joint Training Samples":"30.7M","Parameters (S+V)":"10.9B"}. This record does not assert independent reproduction. ### JavisDiT PQ: 5.21. Result ID: 69c3e40347f34ce924ad6fb1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Generation Paradigm":"Joint Generation","Joint Training Samples":"10.1M","Parameters (S+V)":"3.7B"}. This record does not assert independent reproduction. ### Universe-1 PQ: 4.56. Result ID: 69c3e40347f34ce924ad6fb2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Generation Paradigm":"Joint Generation","Joint Training Samples":"6.4M","Parameters (S+V)":"7.1B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.