arXiv:2509. 09151v2 Announce Type: replace-cross Abstract: Research in video understanding has advanced rapidly, driven by increasingly diverse datasets and more powerful model architectures.
arXiv:2606. 31050v1 Announce Type: cross Abstract: How to accurately predict a high-fidelity future world?
The paper introduces 4DStreamCtrl, a system that unifies camera motion, object trajectories, and depth into a single 3D point‑track representation, enabling joint control, depth editing, and motion transfer in a single forward pass. By mining in‑the‑wild video for 3D motion supervision and encoding it with a lightweight Geometric Motion Head, the authors train a causal streaming student that can generate arbitrarily long videos in just four denoising steps, achieving 20 FPS on a single high‑end GPU for 480p video. This approach outperforms prior camera‑only, 2D, and offline‑3D methods in motion‑control precision while maintaining temporal coherence over hundreds of frames, thereby enabling interactive 4D‑controllable streaming generation for the first time.
SolarWM is an open foundation for building interactive video world models, offering a reconfigurable multi‑source data engine that unifies 1.43 million clips from 10 datasets into a consistent, frame‑aligned format. It provides a backbone‑native adaptation framework that preserves native representations of models ranging from 5 B to 33 B parameters, and a three‑stage training recipe combining bidirectional adaptation, teacher‑forced autoregressive initialization, and distribution‑matching distillation. The resulting causal models can interact in real‑time over rollouts from minutes to hours, trained only on 5‑second sequences, and the project releases data, pipeline, recipes, weights, and framework for reproducible research.
arXiv:2607. 18367v1 Announce Type: new Abstract: Unlike conventional video game development, which relies on labor-intensive pipelines for asset production, animation, physics, and programming, video world models generate interactive environments from user inputs instantly.
Avatar‑Forever introduces a decoupled parallel training framework that separates efficient generation from long‑horizon robustness, allowing each capability to be trained independently. The method employs full‑parameter distillation for high‑quality generation and a lightweight long‑horizon adapter trained via Recovery‑oriented Rollout Training (RRT) to mitigate autoregressive errors. Additionally, a ForeverCache feature‑caching mechanism reduces redundant computation during streaming inference, enabling real‑time, high‑resolution avatar generation at 27.2 FPS on a single H100 GPU.
GenStream is a semantic streaming framework that replaces dense video frames with compact metadata—skeletal keypoints, camera parameters, and a static 3D background model—to enable generative reconstruction of human figures on the client side. By transmitting only structured information rather than full pixel data, it achieves over 99.9% bandwidth reduction compared to HEVC, as demonstrated on Olympic figure skating footage. The approach shifts computational load to the client and opens possibilities for volumetric avatar synthesis, multi‑view actor fusion, and personalized viewing experiences in a post‑codec era.
arXiv:2606. 29531v1 Announce Type: cross Abstract: We propose MotionAtlas, a system for detailed captioning of motion-centric videos, comprising (1) a dedicated human-annotated benchmark, (2) a scalable, high-quality pipeline to construct training samples, and (3) a family of powerful Video-MLLMs.
arXiv:2609.15032v1 Announce Type: new Abstract: Live free-viewpoint visualization of real humans is critical for immersive communication and interactive digital experiences. Existing methods either r...
VideoTIR introduces a reinforcement‑learning approach to improve long‑video understanding by encouraging multimodal large language models to use comprehensive multi‑level toolkits efficiently. It combines Zero‑RL and SFT cold‑starting strategies to help models retrieve and focus on meaningful video segments, images, and regions, thereby reducing hallucinations. The method includes Toolkit Action Grouped Policy Optimization (TAGPO) to streamline tool‑calling and a sandbox‑based trajectory synthesis framework for high‑quality data, achieving strong results on three long‑video QA benchmarks.
arXiv:2604.08995v3 Announce Type: replace Abstract: With the advancement of interactive video generation, diffusion models have increasingly demonstrated their potential as world models. However, exi...
Camera intrinsics are vital for recovering 3D structure from 2D video. However, most 3D algorithms assume fixed intrinsics throughout a video, an assumption that often fails for real-world in-the-wild videos.