OmniVCus: Feedforward Subject-driven Video Customization with Multimodal Control Conditions

• URL: http://arxiv.org/abs/2506.23361v1
• Date: Sun, 29 Jun 2025 18:43:00 GMT
• Title: OmniVCus: Feedforward Subject-driven Video Customization with Multimodal Control Conditions
• Authors: Yuanhao Cai, He Zhang, Xi Chen, Jinbo Xing, Yiwei Hu, Yuqian Zhou, Kai Zhang, Zhifei Zhang, Soo Ye Kim, Tianyu Wang, Yulun Zhang, Xiaokang Yang, Zhe Lin, Alan Yuille,
• Abstract summary: We develop an Image-Video Transfer Mixed (IVTM) training with image editing data to enable instructive editing for the subject in the customized video.<n>We also propose a diffusion Transformer framework, OmniVCus, with two embedding mechanisms, Lottery Embedding (LE) and Temporally Aligned Embedding (TAE)<n>Our method significantly surpasses state-of-the-art methods in both quantitative and qualitative evaluations.
• Score: 96.31455979495398
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Existing feedforward subject-driven video customization methods mainly study single-subject scenarios due to the difficulty of constructing multi-subject training data pairs. Another challenging problem that how to use the signals such as depth, mask, camera, and text prompts to control and edit the subject in the customized video is still less explored. In this paper, we first propose a data construction pipeline, VideoCus-Factory, to produce training data pairs for multi-subject customization from raw videos without labels and control signals such as depth-to-video and mask-to-video pairs. Based on our constructed data, we develop an Image-Video Transfer Mixed (IVTM) training with image editing data to enable instructive editing for the subject in the customized video. Then we propose a diffusion Transformer framework, OmniVCus, with two embedding mechanisms, Lottery Embedding (LE) and Temporally Aligned Embedding (TAE). LE enables inference with more subjects by using the training subjects to activate more frame embeddings. TAE encourages the generation process to extract guidance from temporally aligned control signals by assigning the same frame embeddings to the control and noise tokens. Experiments demonstrate that our method significantly surpasses state-of-the-art methods in both quantitative and qualitative evaluations. Video demos are at our project page: https://caiyuanhao1998.github.io/project/OmniVCus/. Our code will be released at https://github.com/caiyuanhao1998/Open-OmniVCus

Related papers

• InternVideo2: Scaling Foundation Models for Multimodal Video Understanding [51.129913789991924]
  InternVideo2 is a new family of video foundation models (FM) that achieve state-of-the-art results in video recognition, video-speech tasks, and video-centric tasks. Our core design is a progressive training approach that unifies the masked video modeling, cross contrastive learning, and prediction token, scaling up to 6B video size.
  arXiv  Detail & Related papers  (2024-03-22T17:57:42Z)
• Video-LaVIT: Unified Video-Language Pre-training with Decoupled Visual-Motional Tokenization [52.63845811751936]
  Video pre-training is challenging due to the modeling of its dynamics video. In this paper, we address such limitations in video pre-training with an efficient video decomposition. Our framework is both capable of comprehending and generating image and video content, as demonstrated by its performance across 13 multimodal benchmarks.
  arXiv  Detail & Related papers  (2024-02-05T16:30:49Z)
• InstructVid2Vid: Controllable Video Editing with Natural Language Instructions [97.17047888215284]
  InstructVid2Vid is an end-to-end diffusion-based methodology for video editing guided by human language instructions. Our approach empowers video manipulation guided by natural language directives, eliminating the need for per-example fine-tuning or inversion.
  arXiv  Detail & Related papers  (2023-05-21T03:28:13Z)
• Pix2Video: Video Editing using Image Diffusion [43.07444438561277]
  We investigate how to use pre-trained image models for text-guided video editing. Our method works in two simple steps: first, we use a pre-trained structure-guided (e.g., depth) image diffusion model to perform text-guided edits on an anchor frame. We demonstrate that realistic text-guided video edits are possible, without any compute-intensive preprocessing or video-specific finetuning.
  arXiv  Detail & Related papers  (2023-03-22T16:36:10Z)
• AutoTransition: Learning to Recommend Video Transition Effects [20.384463765702417]
  We present the premier work on performing automatic video transitions recommendation (VTR) VTR is given a sequence of raw video shots and companion audio, recommend video transitions for each pair of neighboring shots. We propose a novel multi-modal matching framework which consists of two parts.
  arXiv  Detail & Related papers  (2022-07-27T12:00:42Z)
• Show Me What and Tell Me How: Video Synthesis via Multimodal Conditioning [36.85533835408882]
  This work presents a multimodal video generation framework that benefits from text and images provided jointly or separately. We propose a new video token trained with self-learning and an improved mask-prediction algorithm for sampling video tokens. Our framework can incorporate various visual modalities, such as segmentation masks, drawings, and partially occluded images.
  arXiv  Detail & Related papers  (2022-03-04T21:09:13Z)
• Align and Prompt: Video-and-Language Pre-training with Entity Prompts [111.23364631136339]
  Video-and-language pre-training has shown promising improvements on various downstream tasks. We propose Align and Prompt: an efficient and effective video-and-language pre-training framework with better cross-modal alignment. Our code and pre-trained models will be released.
  arXiv  Detail & Related papers  (2021-12-17T15:55:53Z)
• VIOLET : End-to-End Video-Language Transformers with Masked Visual-token Modeling [88.30109041658618]
  A great challenge in video-language (VidL) modeling lies in the disconnection between fixed video representations extracted from image/video understanding models and downstream VidL data. We present VIOLET, a fully end-to-end VIdeO-LanguagE Transformer, which adopts a video transformer to explicitly model the temporal dynamics of video inputs.
  arXiv  Detail & Related papers  (2021-11-24T18:31:20Z)
• VIMPAC: Video Pre-Training via Masked Token Prediction and Contrastive Learning [82.09856883441044]
  Video understanding relies on perceiving the global content modeling its internal connections. We propose a block-wise strategy where we mask neighboring video tokens in both spatial and temporal domains. We also add an augmentation-free contrastive learning method to further capture global content.
  arXiv  Detail & Related papers  (2021-06-21T16:48:19Z)

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.