HumanDiT: Pose-Guided Diffusion Transformer for Long-form Human Motion Video Generation
This work addresses the limitations of existing human motion video generation methods, particularly for long sequences and intricate motions, which is significant for applications in fields like computer vision and robotics.
The authors tackled the problem of generating high-fidelity, long-form human motion videos with accurate rendering of detailed body parts, achieving superior performance in diverse scenarios. HumanDiT was trained on 14,000 hours of high-quality video and can generate videos with variable sequence lengths and resolutions.
Human motion video generation has advanced significantly, while existing methods still struggle with accurately rendering detailed body parts like hands and faces, especially in long sequences and intricate motions. Current approaches also rely on fixed resolution and struggle to maintain visual consistency. To address these limitations, we propose HumanDiT, a pose-guided Diffusion Transformer (DiT)-based framework trained on a large and wild dataset containing 14,000 hours of high-quality video to produce high-fidelity videos with fine-grained body rendering. Specifically, (i) HumanDiT, built on DiT, supports numerous video resolutions and variable sequence lengths, facilitating learning for long-sequence video generation; (ii) we introduce a prefix-latent reference strategy to maintain personalized characteristics across extended sequences. Furthermore, during inference, HumanDiT leverages Keypoint-DiT to generate subsequent pose sequences, facilitating video continuation from static images or existing videos. It also utilizes a Pose Adapter to enable pose transfer with given sequences. Extensive experiments demonstrate its superior performance in generating long-form, pose-accurate videos across diverse scenarios.