PoseGen: In-Context LoRA Finetuning for Pose-Controllable Long Human Video Generation

arXiv:2508.05091v2 Announce Type: replace Abstract: 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 .

Related

• Screen, Cache, and Match: A Training-Free Causality-Consistent Reference Frame Framework for Human Animation
• Uni-ViGU: Towards Unified Video Generation and Understanding via A Diffusion-Based Video Generator
• InsEdit: Towards Instruction-based Visual Editing via Data-Efficient Video Diffusion Models Adaptation
• Beyond Flicker: Detecting Kinematic Inconsistencies for Generalizable Deepfake Video Detection
• E-3DPSM: A State Machine for Event-Based Egocentric 3D Human Pose Estimation

Source: arXiv cs.CV | 2026-04-13