LoG-VGGT is a memory‑efficient framework for long‑sequence 3D reconstruction that balances local temporal modeling with global camera consistency. It uses cross‑window attention in a small subset of transformer blocks to propagate information across adjacent temporal windows while keeping memory usage bounded. A global camera consistency refinement module further improves long‑horizon pose stability by enforcing scene‑level constraints through cross‑attention between camera and compact register tokens, leading to better depth accuracy and robust camera pose estimation on multiple benchmarks.
By Jingke Zhou, Chenhang Ma, Zhizhou Zhong, Mingkai Liu, Zhuang Zhou, Yicheng ji, Binghua Su, Bo Cai, Xianliang Huang
Streaming autoregressive diffusion models enable real-time, long-horizon video generation, but their training objectives optimize local frame prediction rather than the geometry and dynamics of a coherent world: long rollouts accumulate geometric drift and degrade into static or unnatural motion. Recent bidirectional approaches address this problem using rewards signals built upon 3D Gaussian-Splatting reconstruction.
TKCAM is a new framework for generating controllable camera motion conditioned on text and sparse keyframes, using a 12‑dimensional kinematic representation and a two‑stage masked transformer with residual vector quantization. It allows users to provide free‑form text prompts and RGB observations at selected timestamps, enabling temporally localized visual guidance for coherent trajectory synthesis. The authors also introduce the RealEstate10K‑Cap dataset and a Universal CLaTr Evaluator, showing that TKCAM outperforms recent baselines on FID, text‑motion matching, and retrieval metrics.
By Haozhe Yang, Zhiyang Dou, Zekai Gu, Cheng Lin, Wenping Wang, Yuan Liu, Taku Komura
arXiv:2608. 19556v1 Announce Type: cross Abstract: Streaming autoregressive diffusion models enable real-time, long-horizon video generation, but their training objectives optimize local frame prediction rather than the geometry and dynamics of a coherent world: long rollouts accumulate geometric drift and degrade into static or unnatural motion.
By Yuanhao Ban, Jiaqi Feng, Hengguang Zhou, Xiaohuan Pei, Justin Cui, Cho-Jui Hsieh
BulletTime presents a 4D‑controllable video diffusion framework that separates scene dynamics from camera pose, allowing precise manipulation of both temporal and spatial aspects of generated videos. The model conditions on continuous world‑time sequences and camera trajectories, integrating them via a 4D positional encoding in the attention layer and adaptive normalizations for feature modulation. A specially curated dataset with independently parameterized temporal and camera variations is used for training, and the resulting system demonstrates robust real‑world 4D control while maintaining high generation quality and surpassing prior methods in controllability.
By Yiming Wang, Qihang Zhang, Shengqu Cai, Tong Wu, Jan Ackermann, Zhengfei Kuang, Yang Zheng, Frano Raji\v{c}, Siyu Tang, Gordon Wetzstein
Real-world spatial intelligence requires agents to understand scenes from continuous video streams, where objects move, persist, disappear, and reappear over time. While recent spatial foundation models have enabled generalizable feed-forward 3D reconstruction, most streaming methods remain geometry-centric and lack temporally consistent object-level understanding.