DreamVideo-2: Zero-Shot Subject-Driven Video Customization with Precise Motion Control

• URL: http://arxiv.org/abs/2410.13830v1
• Date: Thu, 17 Oct 2024 17:52:57 GMT
• Title: DreamVideo-2: Zero-Shot Subject-Driven Video Customization with Precise Motion Control
• Authors: Yujie Wei, Shiwei Zhang, Hangjie Yuan, Xiang Wang, Haonan Qiu, Rui Zhao, Yutong Feng, Feng Liu, Zhizhong Huang, Jiaxin Ye, Yingya Zhang, Hongming Shan,
• Abstract summary: We present DreamVideo-2, a zero-shot video customization framework capable of generating videos with a specific subject and motion trajectory. Specifically, we introduce reference attention, which leverages the model's inherent capabilities for subject learning. We devise a mask-guided motion module to achieve precise motion control by fully utilizing the robust motion signal of box masks.
• Score: 42.506988751934685
• License:
• Abstract: Recent advances in customized video generation have enabled users to create videos tailored to both specific subjects and motion trajectories. However, existing methods often require complicated test-time fine-tuning and struggle with balancing subject learning and motion control, limiting their real-world applications. In this paper, we present DreamVideo-2, a zero-shot video customization framework capable of generating videos with a specific subject and motion trajectory, guided by a single image and a bounding box sequence, respectively, and without the need for test-time fine-tuning. Specifically, we introduce reference attention, which leverages the model's inherent capabilities for subject learning, and devise a mask-guided motion module to achieve precise motion control by fully utilizing the robust motion signal of box masks derived from bounding boxes. While these two components achieve their intended functions, we empirically observe that motion control tends to dominate over subject learning. To address this, we propose two key designs: 1) the masked reference attention, which integrates a blended latent mask modeling scheme into reference attention to enhance subject representations at the desired positions, and 2) a reweighted diffusion loss, which differentiates the contributions of regions inside and outside the bounding boxes to ensure a balance between subject and motion control. Extensive experimental results on a newly curated dataset demonstrate that DreamVideo-2 outperforms state-of-the-art methods in both subject customization and motion control. The dataset, code, and models will be made publicly available.

Related papers

• MotionBooth: Motion-Aware Customized Text-to-Video Generation [44.41894050494623]
  MotionBooth is a framework designed for animating customized subjects with precise control over both object and camera movements. We efficiently fine-tune a text-to-video model to capture the object's shape and attributes accurately. Our approach presents subject region loss and video preservation loss to enhance the subject's learning performance.
  arXiv  Detail & Related papers  (2024-06-25T17:42:25Z)
• MotionFollower: Editing Video Motion via Lightweight Score-Guided Diffusion [94.66090422753126]
  MotionFollower is a lightweight score-guided diffusion model for video motion editing. It delivers superior motion editing performance and exclusively supports large camera movements and actions. Compared with MotionEditor, the most advanced motion editing model, MotionFollower achieves an approximately 80% reduction in GPU memory.
  arXiv  Detail & Related papers  (2024-05-30T17:57:30Z)
• Video Diffusion Models are Training-free Motion Interpreter and Controller [20.361790608772157]
  This paper introduces a novel perspective to understand, localize, and manipulate motion-aware features in video diffusion models. We present a new MOtion FeaTure (MOFT) by eliminating content correlation information and filtering motion channels.
  arXiv  Detail & Related papers  (2024-05-23T17:59:40Z)
• Investigating the Effectiveness of Cross-Attention to Unlock Zero-Shot Editing of Text-to-Video Diffusion Models [52.28245595257831]
  Cross-attention guidance can be a promising approach for editing videos. We show that despite the limitations of current T2V models, cross-attention guidance can be a promising approach for editing videos.
  arXiv  Detail & Related papers  (2024-04-08T13:40:01Z)
• Animate Your Motion: Turning Still Images into Dynamic Videos [58.63109848837741]
  We introduce Scene and Motion Conditional Diffusion (SMCD), a novel methodology for managing multimodal inputs. SMCD incorporates a recognized motion conditioning module and investigates various approaches to integrate scene conditions. Our design significantly enhances video quality, motion precision, and semantic coherence.
  arXiv  Detail & Related papers  (2024-03-15T10:36:24Z)
• TrailBlazer: Trajectory Control for Diffusion-Based Video Generation [11.655256653219604]
  Controllability in text-to-video (T2V) generation is often a challenge. We introduce the concept of keyframing, allowing the subject trajectory and overall appearance to be guided by both a moving bounding box and corresponding prompts. Despite the simplicity of the bounding box guidance, the resulting motion is surprisingly natural, with emergent effects including perspective and movement toward the virtual camera as the box size increases.
  arXiv  Detail & Related papers  (2023-12-31T10:51:52Z)
• DreamVideo: Composing Your Dream Videos with Customized Subject and Motion [52.7394517692186]
  We present DreamVideo, a novel approach to generating personalized videos from a few static images of the desired subject. DreamVideo decouples this task into two stages, subject learning and motion learning, by leveraging a pre-trained video diffusion model. In motion learning, we architect a motion adapter and fine-tune it on the given videos to effectively model the target motion pattern.
  arXiv  Detail & Related papers  (2023-12-07T16:57:26Z)
• TrackDiffusion: Tracklet-Conditioned Video Generation via Diffusion Models [75.20168902300166]
  We propose TrackDiffusion, a novel video generation framework affording fine-grained trajectory-conditioned motion control. A pivotal component of TrackDiffusion is the instance enhancer, which explicitly ensures inter-frame consistency of multiple objects. generated video sequences by our TrackDiffusion can be used as training data for visual perception models.
  arXiv  Detail & Related papers  (2023-12-01T15:24:38Z)

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.