Generating, Fast and Slow: Scalable Parallel Video Generation with Video Interface Networks

• URL: http://arxiv.org/abs/2503.17539v1
• Date: Fri, 21 Mar 2025 21:13:02 GMT
• Title: Generating, Fast and Slow: Scalable Parallel Video Generation with Video Interface Networks
• Authors: Bhishma Dedhia, David Bourgin, Krishna Kumar Singh, Yuheng Li, Yan Kang, Zhan Xu, Niraj K. Jha, Yuchen Liu,
• Abstract summary: We introduce a new paradigm called Video Interface Networks (VINs), which augment DiTs with an abstraction module to enable parallel inference of video chunks.<n>VINs encode global semantics from the noisy input of local chunks and the encoded representations, in turn, guide DiTs in denoising chunks in parallel.<n>Our approach attains state-of-the-art motion smoothness while using 25-40% fewer FLOPs than full generation.
• Score: 21.710127132217526
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Diffusion Transformers (DiTs) can generate short photorealistic videos, yet directly training and sampling longer videos with full attention across the video remains computationally challenging. Alternative methods break long videos down into sequential generation of short video segments, requiring multiple sampling chain iterations and specialized consistency modules. To overcome these challenges, we introduce a new paradigm called Video Interface Networks (VINs), which augment DiTs with an abstraction module to enable parallel inference of video chunks. At each diffusion step, VINs encode global semantics from the noisy input of local chunks and the encoded representations, in turn, guide DiTs in denoising chunks in parallel. The coupling of VIN and DiT is learned end-to-end on the denoising objective. Further, the VIN architecture maintains fixed-size encoding tokens that encode the input via a single cross-attention step. Disentangling the encoding tokens from the input thus enables VIN to scale to long videos and learn essential semantics. Experiments on VBench demonstrate that VINs surpass existing chunk-based methods in preserving background consistency and subject coherence. We then show via an optical flow analysis that our approach attains state-of-the-art motion smoothness while using 25-40% fewer FLOPs than full generation. Finally, human raters favorably assessed the overall video quality and temporal consistency of our method in a user study.

Related papers

• Token-Efficient Long Video Understanding for Multimodal LLMs [101.70681093383365]
  STORM is a novel architecture incorporating a dedicated temporal encoder between the image encoder and the Video-LLMs.<n>We show that STORM achieves state-of-the-art results across various long video understanding benchmarks.
  arXiv  Detail & Related papers  (2025-03-06T06:17:38Z)
• DiTCtrl: Exploring Attention Control in Multi-Modal Diffusion Transformer for Tuning-Free Multi-Prompt Longer Video Generation [54.30327187663316]
  DiTCtrl is a training-free multi-prompt video generation method under MM-DiT architectures for the first time. We analyze MM-DiT's attention mechanism, finding that the 3D full attention behaves similarly to that of the cross/self-attention blocks in the UNet-like diffusion models. Based on our careful design, the video generated by DiTCtrl achieves smooth transitions and consistent object motion given multiple sequential prompts.
  arXiv  Detail & Related papers  (2024-12-24T18:51:19Z)
• SF-V: Single Forward Video Generation Model [57.292575082410785]
  We propose a novel approach to obtain single-step video generation models by leveraging adversarial training to fine-tune pre-trained models. Experiments demonstrate that our method achieves competitive generation quality of synthesized videos with significantly reduced computational overhead.
  arXiv  Detail & Related papers  (2024-06-06T17:58:27Z)
• 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)
• TAM-VT: Transformation-Aware Multi-scale Video Transformer for Segmentation and Tracking [33.75267864844047]
  Video Object (VOS) has emerged as an increasingly important problem with availability of larger datasets and more complex and realistic settings. We propose a novel, clip-based DETR-style encoder-decoder architecture, which focuses on systematically analyzing and addressing aforementioned challenges. Specifically, we propose a novel transformation-aware loss that focuses learning on portions of the video where an object undergoes significant deformations.
  arXiv  Detail & Related papers  (2023-12-13T21:02:03Z)
• MED-VT++: Unifying Multimodal Learning with a Multiscale Encoder-Decoder Video Transformer [12.544216587327387]
  We present an end-to-end trainable unified multiscale encoder-decoder transformer that is focused on dense prediction tasks in video. The presented Multiscale-Decoder Video (MED-VT) uses multiscale representation throughout and employs an optional input beyond video. We present a transductive learning scheme through many-to-many label propagation to provide temporally consistent video predictions.
  arXiv  Detail & Related papers  (2023-04-12T15:50:19Z)
• Learning Trajectory-Aware Transformer for Video Super-Resolution [50.49396123016185]
  Video super-resolution aims to restore a sequence of high-resolution (HR) frames from their low-resolution (LR) counterparts. Existing approaches usually align and aggregate video frames from limited adjacent frames. We propose a novel Transformer for Video Super-Resolution (TTVSR)
  arXiv  Detail & Related papers  (2022-04-08T03:37:39Z)
• 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)
• Multi-Density Attention Network for Loop Filtering in Video Compression [9.322800480045336]
  We propose a on-line scaling based multi-density attention network for loop filtering in video compression. Experimental results show that 10.18% bit-rate reduction at the same video quality can be achieved over the latest Versatile Video Coding (VVC) standard.
  arXiv  Detail & Related papers  (2021-04-08T05:46:38Z)

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.