# Text-to-Motion Generation on KeyframeFace (test) Source: [Text-to-Motion Generation on KeyframeFace (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:44:40.585Z Evaluation of models generating motion sequences conditioned on text prompts using the KeyframeFace test set. Metrics include FID, W-Dist, R-Precision variants, Diversity, and MMD. 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 | Text-to-Motion Generation | | Dataset (including any stated split) | KeyframeFace (test) | | 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](): Lower is better; unit: SCALAR. - [W-Dist](): Lower is better; unit: SCALAR. - [R-Precision (top-1)](): Higher is better; unit: PERCENT. - [R-Precision (top-2)](): Higher is better; unit: PERCENT. - [R-Precision (top-3)](): Higher is better; unit: PERCENT. - [Diversity](): Higher is better; unit: SCALAR. - [MMD](): Lower is better; unit: SCALAR. ## Recorded results | Method | FID | Result date | Report paper | | --- | --- | --- | --- | | Express4D-MDM | 0.0348 | Not recorded | 693f95dc6cfd488761a817fd | | Tuned-semantic-14B | 0.0599 | Not recorded | 693f95dc6cfd488761a817fd | | Tuned-semantic-4B | 0.0777 | Not recorded | 693f95dc6cfd488761a817fd | | Tuned-non-semantic-4B | 0.2406 | Not recorded | 693f95dc6cfd488761a817fd | | Tuned-non-semantic-14B | 0.2805 | Not recorded | 693f95dc6cfd488761a817fd | | Base-semantic-4B | 0.8557 | Not recorded | 693f95dc6cfd488761a817fd | | Base-non-semantic-4B | 4.3623 | Not recorded | 693f95dc6cfd488761a817fd | ### Express4D-MDM FID: 0.0348. Result ID: 6998571b5698665ec4b5254c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Architecture":"Transformer-based denoising diffusion","Dataset":"KeyframeFace (adapted from Express4D)"}. This record does not assert independent reproduction. ### Tuned-semantic-14B FID: 0.0599. Result ID: 6998571b5698665ec4b5254e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Representation":"Semantic (P\_sem)","Training Status":"Fine-tuned (LoRA)","Scale":"14B","Backbone":"Qwen3-14B / DeepSeek-R1-Distill-Qwen-14B"}. This record does not assert independent reproduction. ### Tuned-semantic-4B FID: 0.0777. Result ID: 6998571b5698665ec4b5254b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Representation":"Semantic (P\_sem)","Training Status":"Fine-tuned (LoRA)","Scale":"4B","Backbone":"Qwen3-4B-Instruct"}. This record does not assert independent reproduction. ### Tuned-non-semantic-4B FID: 0.2406. Result ID: 6998571b5698665ec4b5254a. [Reporting paper]() Recorded evaluation configuration: {"Representation":"Numerical (non-semantic)","Training Status":"Fine-tuned (LoRA)","Scale":"4B","Backbone":"Qwen3-4B-Instruct"}. This record does not assert independent reproduction. ### Tuned-non-semantic-14B FID: 0.2805. Result ID: 6998571b5698665ec4b5254d. [Reporting paper]() Recorded evaluation configuration: {"Representation":"Numerical (non-semantic)","Training Status":"Fine-tuned (LoRA)","Scale":"14B","Backbone":"Qwen3-14B / DeepSeek-R1-Distill-Qwen-14B"}. This record does not assert independent reproduction. ### Base-semantic-4B FID: 0.8557. Result ID: 6998571b5698665ec4b52549. [Reporting paper]() Recorded evaluation configuration: {"Representation":"Semantic (P\_sem)","Training Status":"Pre-trained (Base)","Scale":"4B","Backbone":"Qwen3-4B-Instruct"}. This record does not assert independent reproduction. ### Base-non-semantic-4B FID: 4.3623. Result ID: 6998571b5698665ec4b52548. [Reporting paper]() Recorded evaluation configuration: {"Representation":"Numerical (non-semantic)","Training Status":"Pre-trained (Base)","Scale":"4B","Backbone":"Qwen3-4B-Instruct"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.