MotionStream: Real-Time Video Generation with Interactive Motion Controls

• URL: http://arxiv.org/abs/2511.01266v1
• Date: Mon, 03 Nov 2025 06:37:53 GMT
• Title: MotionStream: Real-Time Video Generation with Interactive Motion Controls
• Authors: Joonghyuk Shin, Zhengqi Li, Richard Zhang, Jun-Yan Zhu, Jaesik Park, Eli Schechtman, Xun Huang,
• Abstract summary: We present MotionStream, enabling sub-second latency with up to 29 FPS streaming generation on a single GPU.<n>Our approach begins by augmenting a text-to-video model with motion control, which generates high-quality videos that adhere to the global text prompt and local motion guidance, but does not perform inference on the fly.<n>Our models achieve state-of-the-art results in motion following and video quality while being two orders of magnitude faster, uniquely enabling infinite-length streaming.
• Score: 60.403597895657505
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Current motion-conditioned video generation methods suffer from prohibitive latency (minutes per video) and non-causal processing that prevents real-time interaction. We present MotionStream, enabling sub-second latency with up to 29 FPS streaming generation on a single GPU. Our approach begins by augmenting a text-to-video model with motion control, which generates high-quality videos that adhere to the global text prompt and local motion guidance, but does not perform inference on the fly. As such, we distill this bidirectional teacher into a causal student through Self Forcing with Distribution Matching Distillation, enabling real-time streaming inference. Several key challenges arise when generating videos of long, potentially infinite time-horizons: (1) bridging the domain gap from training on finite length and extrapolating to infinite horizons, (2) sustaining high quality by preventing error accumulation, and (3) maintaining fast inference, without incurring growth in computational cost due to increasing context windows. A key to our approach is introducing carefully designed sliding-window causal attention, combined with attention sinks. By incorporating self-rollout with attention sinks and KV cache rolling during training, we properly simulate inference-time extrapolations with a fixed context window, enabling constant-speed generation of arbitrarily long videos. Our models achieve state-of-the-art results in motion following and video quality while being two orders of magnitude faster, uniquely enabling infinite-length streaming. With MotionStream, users can paint trajectories, control cameras, or transfer motion, and see results unfold in real-time, delivering a truly interactive experience.

Related papers

• Knot Forcing: Taming Autoregressive Video Diffusion Models for Real-time Infinite Interactive Portrait Animation [16.692450893925148]
  We present a novel streaming framework named Knot Forcing for real-time portrait animation.<n>K Knot Forcing enables high-fidelity, temporally consistent, and interactive portrait animation over infinite sequences.
  arXiv  Detail & Related papers  (2025-12-25T16:34:56Z)
• Deep Forcing: Training-Free Long Video Generation with Deep Sink and Participative Compression [36.99018442740971]
  We find that naively applying StreamingLLM-style attention sinks to video diffusion leads to fidelity degradation and motion stagnation.<n>We introduce Deep Forcing, which consists of two training-free mechanisms that address this without any fine-tuning.<n>Our results demonstrate that training-free KV-cache management can match or exceed training-based approaches for autoregressively streaming long-video generation.
  arXiv  Detail & Related papers  (2025-12-04T18:46:44Z)
• Rolling Forcing: Autoregressive Long Video Diffusion in Real Time [86.40480237741609]
  Rolling Forcing is a novel video generation technique that enables streaming long videos with minimal error accumulation.<n>Rolling Forcing comes with three novel designs. First, instead of iteratively sampling individual frames, which accelerates error propagation, we design a joint denoising scheme.<n>Second, we introduce the attention sink mechanism into the long-horizon stream video generation task, which allows the model to keep key value states of initial frames as a global context anchor.<n>Third, we design an efficient training algorithm that enables few-step distillation over largely extended denoising windows.
  arXiv  Detail & Related papers  (2025-09-29T17:57:14Z)
• Training-Free Efficient Video Generation via Dynamic Token Carving [54.52061549312799]
  Jenga is an inference pipeline that combines dynamic attention carving with progressive resolution generation.<n>As a plug-and-play solution, Jenga enables practical, high-quality video generation on modern hardware.
  arXiv  Detail & Related papers  (2025-05-22T16:21:32Z)
• Learning Streaming Video Representation via Multitask Training [38.83730146750222]
  Unlike offline video understanding, streaming video understanding requires the ability to process video streams frame by frame, preserve historical information, and make low-latency decisions.<n>We develop a novel streaming video backbone, termed as StreamFormer, by incorporating causal temporal attention into a pre-trained vision transformer.<n>We conduct extensive experiments on online action detection, online video instance segmentation, and video question answering.
  arXiv  Detail & Related papers  (2025-04-28T17:59:54Z)
• StreamMind: Unlocking Full Frame Rate Streaming Video Dialogue through Event-Gated Cognition [20.608124640950276]
  We introduce StreamMind, a video LLM framework that achieves ultra-FPS streaming video processing (100 fps on a single A100)<n>We propose a novel perception-cognition intertemporal paradigm named ''event-gated LLM invocation''<n> Experiments on Ego4D and SoccerNet streaming tasks, as well as standard offline benchmarks, demonstrate state-of-the-art performance in both model capability and real-time efficiency.
  arXiv  Detail & Related papers  (2025-03-08T13:44:38Z)
• From Slow Bidirectional to Fast Autoregressive Video Diffusion Models [48.35054927704544]
  Current video diffusion models achieve impressive generation quality but struggle in interactive applications due to bidirectional attention dependencies.<n>We address this limitation by adapting a pretrained bidirectional diffusion transformer to an autoregressive transformer that generates frames on-the-fly.<n>Our model achieves a total score of 84.27 on the VBench-Long benchmark, surpassing all previous video generation models.
  arXiv  Detail & Related papers  (2024-12-10T18:59:50Z)
• Motion Prompting: Controlling Video Generation with Motion Trajectories [57.049252242807874]
  We train a video generation model conditioned on sparse or dense video trajectories.<n>We translate high-level user requests into detailed, semi-dense motion prompts.<n>We demonstrate our approach through various applications, including camera and object motion control, "interacting" with an image, motion transfer, and image editing.
  arXiv  Detail & Related papers  (2024-12-03T18:59:56Z)
• Live2Diff: Live Stream Translation via Uni-directional Attention in Video Diffusion Models [64.2445487645478]
  Large Language Models have shown remarkable efficacy in generating streaming data such as text and audio. We present Live2Diff, the first attempt at designing a video diffusion model with uni-directional temporal attention, specifically targeting live streaming video translation.
  arXiv  Detail & Related papers  (2024-07-11T17:34:51Z)
• ControlVideo: Training-free Controllable Text-to-Video Generation [117.06302461557044]
  ControlVideo is a framework to enable natural and efficient text-to-video generation. It generates both short and long videos within several minutes using one NVIDIA 2080Ti.
  arXiv  Detail & Related papers  (2023-05-22T14:48:53Z)
• Real-time Streaming Video Denoising with Bidirectional Buffers [48.57108807146537]
  Real-time denoising algorithms are typically adopted on the user device to remove the noise involved during the shooting and transmission of video streams. Recent multi-output inference works propagate the bidirectional temporal feature with a parallel or recurrent framework. We propose a Bidirectional Streaming Video Denoising framework, to achieve high-fidelity real-time denoising for streaming videos with both past and future temporal receptive fields.
  arXiv  Detail & Related papers  (2022-07-14T14:01:03Z)

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.