HumanDiT: Pose-Guided Diffusion Transformer for Long-form Human Motion Video Generation

• URL: http://arxiv.org/abs/2502.04847v2
• Date: Mon, 10 Feb 2025 14:51:29 GMT
• Title: HumanDiT: Pose-Guided Diffusion Transformer for Long-form Human Motion Video Generation
• Authors: Qijun Gan, Yi Ren, Chen Zhang, Zhenhui Ye, Pan Xie, Xiang Yin, Zehuan Yuan, Bingyue Peng, Jianke Zhu,
• Abstract summary: We propose HumanDiT, a pose-guided Diffusion Transformer (DiT)-based framework trained on a dataset containing 14,000 hours of high-quality video. HumanDiT supports numerous video resolutions and variable sequence lengths, facilitating learning for long-sequence video generation. Experiments demonstrate its superior performance in generating long-form, pose-accurate videos across diverse scenarios.
• Score: 39.69554411714128
• License:
• Abstract: 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.

Related papers

• FreeLong: Training-Free Long Video Generation with SpectralBlend Temporal Attention [57.651429116402554]
  This paper investigates a straightforward and training-free approach to extend an existing short video diffusion model for consistent long video generation. We find that directly applying the short video diffusion model to generate long videos can lead to severe video quality degradation. Motivated by this, we propose a novel solution named FreeLong to balance the frequency distribution of long video features during the denoising process.
  arXiv  Detail & Related papers  (2024-07-29T11:52:07Z)
• Anchored Diffusion for Video Face Reenactment [17.343307538702238]
  We introduce Anchored Diffusion, a novel method for synthesizing relatively long and seamless videos. We train our model on video sequences with random non-uniform temporal spacing and incorporate temporal information via external guidance. During inference, we leverage the transformer architecture to modify the diffusion process, generating a batch of non-uniform sequences anchored to a common frame.
  arXiv  Detail & Related papers  (2024-07-21T13:14:17Z)
• WildVidFit: Video Virtual Try-On in the Wild via Image-Based Controlled Diffusion Models [132.77237314239025]
  Video virtual try-on aims to generate realistic sequences that maintain garment identity and adapt to a person's pose and body shape in source videos. Traditional image-based methods, relying on warping and blending, struggle with complex human movements and occlusions. We reconceptualize video try-on as a process of generating videos conditioned on garment descriptions and human motion. Our solution, WildVidFit, employs image-based controlled diffusion models for a streamlined, one-stage approach.
  arXiv  Detail & Related papers  (2024-07-15T11:21:03Z)
• MultiPly: Reconstruction of Multiple People from Monocular Video in the Wild [32.6521941706907]
  We present MultiPly, a novel framework to reconstruct multiple people in 3D from monocular in-the-wild videos. We first define a layered neural representation for the entire scene, composited by individual human and background models. We learn the layered neural representation from videos via our layer-wise differentiable volume rendering.
  arXiv  Detail & Related papers  (2024-06-03T17:59:57Z)
• EasyAnimate: A High-Performance Long Video Generation Method based on Transformer Architecture [11.587428534308945]
  EasyAnimate is an advanced method for video generation that leverages the power of transformer architecture for high-performance outcomes. We have expanded the DiT framework originally designed for 2D image synthesis to accommodate the complexities of 3D video generation by incorporating a motion module block. We provide a holistic ecosystem for video production based on DiT, encompassing aspects such as data pre-processing, VAE training, DiT models training, and end-to-end video inference.
  arXiv  Detail & Related papers  (2024-05-29T11:11:07Z)
• VITON-DiT: Learning In-the-Wild Video Try-On from Human Dance Videos via Diffusion Transformers [53.45587477621942]
  We propose the first DiT-based video try-on framework for practical in-the-wild applications, named VITON-DiT. Specifically, VITON-DiT consists of a garment extractor, a Spatial-Temporal denoising DiT, and an identity preservation ControlNet. We also introduce random selection strategies during training and an Interpolated Auto-Regressive (IAR) technique at inference to facilitate long video generation.
  arXiv  Detail & Related papers  (2024-05-28T16:21:03Z)
• VividPose: Advancing Stable Video Diffusion for Realistic Human Image Animation [79.99551055245071]
  We propose VividPose, an end-to-end pipeline that ensures superior temporal stability. An identity-aware appearance controller integrates additional facial information without compromising other appearance details. A geometry-aware pose controller utilizes both dense rendering maps from SMPL-X and sparse skeleton maps. VividPose exhibits superior generalization capabilities on our proposed in-the-wild dataset.
  arXiv  Detail & Related papers  (2024-05-28T13:18:32Z)
• MAVIN: Multi-Action Video Generation with Diffusion Models via Transition Video Infilling [19.004339956475498]
  MAVIN is designed to generate transition videos that seamlessly connect two given videos, forming a cohesive integrated sequence. We introduce a new metric, CLIP-RS (CLIP Relative Smoothness), to evaluate temporal coherence and smoothness, complementing traditional quality-based metrics. Experimental results on horse and tiger scenarios demonstrate MAVIN's superior performance in generating smooth and coherent video transitions.
  arXiv  Detail & Related papers  (2024-05-28T09:46:09Z)
• Generating Long Videos of Dynamic Scenes [66.56925105992472]
  We present a video generation model that reproduces object motion, changes in camera viewpoint, and new content that arises over time. A common failure case is for content to never change due to over-reliance on inductive biases to provide temporal consistency.
  arXiv  Detail & Related papers  (2022-06-07T16:29:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.