arXiv:2608.27529v1 Announce Type: new Abstract: Streaming 3D reconstruction from extremely long videos requires estimating camera motion and scene geometry online under bounded memory and computation...
Streaming 3D reconstruction relies on a compact recurrent scene state to process long image streams in linear time and bounded memory. However, repeated updates can gradually corrupt this state, causing reliable historical information to be overwritten by noisy or ambiguous observations.
Anchor3R is a streaming 3D reconstruction framework that predicts window-relative poses and local geometry in the current‑frame coordinate system, forming a dense relative‑pose graph for online pose updates and loop‑aware motion averaging. It improves long‑horizon pose accuracy and dense reconstruction quality on indoor, outdoor, driving, and RGB‑D benchmarks, and generalizes from 48‑frame training sequences to streams exceeding 10,000 frames while keeping GPU memory bounded. The method addresses issues of train‑test mismatch, early‑anchor bias, and accumulated drift found in previous streaming models.
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.
Feed-forward 3D Gaussian Splatting enables efficient novel-view synthesis without per-scene optimization, but most existing methods assume a fixed set of context views and process them jointly. This limits their applicability to online scenarios where calibrated views arrive sequentially and the scene must be updated causally.
arXiv:2603.01765v5 Announce Type: replace Abstract: Monocular depth foundation models generalize across diverse scenes, but recovering accurate metric depth consistent with a target sensor remains ch...
LongVU‑TTT is a causal test‑time training method for long‑video multimodal large language models that inserts a convolutional resampler with fast‑weight updates between the vision encoder and the LLM. The fast weights adapt per video and contextualize frame features before compression, while a hybrid selector keeps explicit visual evidence for downstream reasoning. Experiments show that TTT‑Conv outperforms TTT‑MLP and bidirectional Mamba2 on MLVU, and beats attention‑ and fixed‑state recurrent resamplers on three benchmarks, achieving competitive results on five video‑understanding tasks after reducing 512 frames to 128 LLM frames.
arXiv:2609.23286v1 Announce Type: new Abstract: 3D reconstruction from a lengthy video stream input poses a dilemma for feed-forward reconstruction models (FFRMs), that a whole-stream inference conte...
arXiv:2512. 02473v2 Announce Type: replace-cross Abstract: Video world models have attracted significant attention for their ability to produce high-fidelity future visual observations conditioned on past observations and navigation actions.
TT-VidT is a video pretraining method that decouples the temporal axis by combining a per‑frame ViT-B/16 spatial encoder with a compact Temporal Transfer Layer trained via Diff Compression. The authors conduct a systematic 24‑configuration study to isolate architecture, objective, and decoder effects, showing that the full TT-VidT design yields the strongest motion‑sensitive representations. In downstream fine‑tuning, TT‑VidT outperforms state‑of‑the‑art baselines on Jester, Something‑Something V2, ARID, and Diving48 while using significantly fewer encoder FLOPs.
arXiv:2606.06158v2 Announce Type: replace Abstract: Adaptive video tokenisation seeks to dynamically allocate token budgets based on the underlying visual complexity of a sequence. Current continuous...
Long-video MLLMs must model temporal change before a limited visual-token budget removes most frame evidence. We introduce LongVU-TTT, which inserts a convolutional Test-Time Training (TTT) resampler...