Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2508.05091 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913020173615104 |
|---|---|
| author | He, Jingxuan Su, Busheng Wong, Finn |
| author_facet | He, Jingxuan Su, Busheng Wong, Finn |
| contents | Generating temporally coherent, long-duration videos with precise control over subject identity and movement remains a fundamental challenge for contemporary diffusion-based models, which often suffer from identity drift and are limited to short video length. We present PoseGen, a novel framework that generates human videos of extended duration from a single reference image and a driving video. Our contributions include an in-context LoRA finetuning design that injects subject appearance at the token level for identity preservation, while simultaneously conditioning on pose information at the channel level for fine-grained motion control. To overcome duration limits, we introduce a segment-interleaved generation strategy, where non-overlapping segments are first generated with improved background consistency through a shared KV-cache mechanism, and then stitched into a continuous sequence via pose-aware interpolated generation. Despite being trained on a remarkably small 33-hour video dataset, PoseGen demonstrates superior performance over state-of-the-art baselines in identity fidelity, pose accuracy, and temporal consistency. Code is available at https://github.com/Jessie459/PoseGen . |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_05091 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | PoseGen: In-Context LoRA Finetuning for Pose-Controllable Long Human Video Generation He, Jingxuan Su, Busheng Wong, Finn Computer Vision and Pattern Recognition Generating temporally coherent, long-duration videos with precise control over subject identity and movement remains a fundamental challenge for contemporary diffusion-based models, which often suffer from identity drift and are limited to short video length. We present PoseGen, a novel framework that generates human videos of extended duration from a single reference image and a driving video. Our contributions include an in-context LoRA finetuning design that injects subject appearance at the token level for identity preservation, while simultaneously conditioning on pose information at the channel level for fine-grained motion control. To overcome duration limits, we introduce a segment-interleaved generation strategy, where non-overlapping segments are first generated with improved background consistency through a shared KV-cache mechanism, and then stitched into a continuous sequence via pose-aware interpolated generation. Despite being trained on a remarkably small 33-hour video dataset, PoseGen demonstrates superior performance over state-of-the-art baselines in identity fidelity, pose accuracy, and temporal consistency. Code is available at https://github.com/Jessie459/PoseGen . |
| title | PoseGen: In-Context LoRA Finetuning for Pose-Controllable Long Human Video Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2508.05091 |