从 Kinetics 到 Video-MME 的基准演进、视频编码器到 VLM 的模型迭代、token 瓶颈与帧采样难题：多模态视频理解的进展全景与开放问题。

Browser Use 开源的会话式视频剪辑技能包 video-use（当日涨星 +745、★26,450、MIT）：不让模型看像素，只让它读 12KB 词级转写文本 + 按需抽查画面，配合 12 条硬规则与 EDL 驱动的 ffmpeg 确定性渲染管线。拆解两条读入通道、六步流水线与自评回路，附五个落地场景与成本口径。

Few-step autoregressive video diffusion enables efficient generation, yet maintaining consistency over long rollouts rem…

Multimodal large language models (MLLMs) can generate detailed video captions, yet they may overemphasize minor details …

Prediction errors from video world models are often treated as physical surprise, but they can also reflect appearance, …

Real-time long-video generation requires continuous output with consistent appearance and natural motion. Existing chunk…

Predicting future physical states stands as a central challenge in embodied intelligence. Existing video world models le…

Video generation is shifting from end-to-end synthesis toward agentic generation, where planning, tool use, and feedback…

Oct 25, 2026 Nvidia Labeling one hour of robot video for 72 cents instead of $88 is what Nvidia promises with Flora, its…

Cinematic video generation requires both temporal inference from limited inputs (i.e., sparse rather than frame-by-frame…

Code-based video generation offers fine-grained control, targeted editing, and component reuse compared with direct vide…

Video scene graph generation (VidSGG) aims to capture objects and their evolving relationships in videos. This requires …

在 YouTube 获得美国奥斯卡奖颁奖典礼的转播权之后，另一个流媒体平台亚马逊 Prime Video 获得了美国另一个主要奖项艾美奖的转播权。Amazon Prime Video 将从 2027 年起成为艾美奖的全球独占播放平台，这一协…

Data video is a media form that integrates data visualization with video narrative, widely adopted in news reporting and…

Single-image, multi-image, and video deep research require different visual operations but share a workflow of visual gr…

A 4D audio-visual scene comprises a video, the dynamic geometry it depicts, and the sound sources that populate it, each…

Video generators and video-based world models can synthesize plausible visual trajectories, but long-horizon procedural …

Video-language models commonly assume that more temporal observations lead to more reliable reasoning. We question this …

Recovering faithful 3D hand motion from video remains challenging due to frequent occlusions and incomplete visual obser…

Interactive video world models need to generate each video chunk efficiently while responding faithfully to user control…

Bidirectional video diffusion models denoise entire videos in parallel, yet when trained on effectively unlimited in-dis…

Autoregressive video diffusion supports streaming generation and interactive control, but its KV cache grows continuousl…

Recent video generation models can produce highly realistic videos from natural language instructions, with visual quali…

Streaming video LLMs must retain evidence before its relevance to future tasks is known and respond when sufficient evid…

Recent video-based world models pair the scalability of autoregressive (AR) prediction with the visual quality of diffus…

Video world models require persistent scene memory to maintain consistency during long-horizon video generation. Existin…

Video generation has rapidly progressed from short, low-quality clips to high-resolution, long-duration sequences with c…

Video understanding agents acquire evidence through an executable harness that controls what they observe and how they u…

Google Research has introduced an **AI video co-director** for long-form video generation. The suite of 4 agentic framew…

World-space hand motion estimation from egocentric video requires recovering 3D articulated hand geometry while tracking…

Long-horizon camera-controlled video generation requires recovering previously observed content from an ever-growing vis…

Comparisons in video self-supervised learning often evaluate complete training recipes rather than isolating the method …

Few-step autoregressive video diffusion generates a long video by splitting the video into temporal chunks and generatin…

Streaming video understanding requires models to interpret evidence as it arrives, yet current evaluations often report …

Few-step autoregressive (AR) video diffusion enables low-latency streaming generation, but existing post-training method…

Recent years have witnessed major progress in joint audio-video generation. Existing models still suffer from limited pe…

Video generation begins in text space by authoring a cinematic screenplay, then materializes into pixels. As contemporar…

Advances in generative models have improved video fidelity, enabling long-horizon generation, interactive world modeling…

Video is a rich representation of a physical event, capturing appearance, geometry, motion, and temporal evolution. Othe…

Recent advances in video generative models have enabled high-fidelity, temporally coherent video generation. However, th…

In this paper, we propose SVEET, a framework that requires merely training on a pretrained bidirectional video diffusion…

Video world models enable interactive exploration of dynamic environments, yet struggle to respect prior observations ov…

Runway wants to stream AI video as users prompt it, rather than make them wait for finished clips. The approach builds o…

Despite impressive visual quality, state-of-the-art video diffusion models often generate content that violates real-wor…

Reinforcement learning (RL) is vital for optimizing video generation models, with a robust reward model (RM) serving as …

Alibaba's Qwen3.8-Omni-Flash understands audio and video, plans tasks, calls tools, and reports about 45.7% fewer tokens…

Reference-to-video (R2V) generation is evolving toward increasingly general and versatile reference control, giving rise…

Video diffusion models repeatedly process long spatiotemporal token sequences during denoising, making attention a major…

Nunchux AI has released VC-Attention, a training-free low-bit attention kernel built for video Diffusion Transformers (D…

Interactive control for video generation is moving from coarse prompts toward fine-grained, physically meaningful manipu…