We study relative position encoding for multi-view vision Transformers under camera heterogeneity, including varying fields of view (FoVs) or projection models. Existing rotary relative position encod...
The paper introduces G-ray, a ray-level relative position encoding for multi-view vision Transformers that remains consistent across different camera projections. By parameterizing rotary phases with camera-local ray angles, G-ray achieves projection-invariant positional consistency and can be integrated with existing encodings without extra learned parameters. Experiments on 3D reconstruction and novel-view synthesis benchmarks show that G-ray improves performance, notably reducing mean pointmap relative error by 45.8% over MapAnything.
By Shuo Zhang, Xin Su, Wei Wang, Jun Liu, Xinrui Zeng, Yongsen Chen, Chenjie Wang, Guibo Zhu, Jinqiao Wang, Bin Luo, Liangpei Zhang
arXiv:2605.12938v2 Announce Type: replace-cross
Abstract: Video world models should predict future appearance in a way that remains consistent with 3D scene structure, camera motion, and lens geometr...
By Seonghyun Jin, Youngmin Kim, Sunwoo Park, Jong Chul Ye
The paper proposes a new sequence-to-sequence formulation for multi-view stereo (MVS) that jointly predicts 3D geometry for all input views using a global transformer architecture. It introduces ray‑map embeddings to inject camera parameters into image tokens and a unified global cost volume to capture 3D structure across all views. Experiments on public benchmarks demonstrate state‑of‑the‑art performance, outperforming both traditional MVS and feed‑forward reconstruction baselines.
By Aoxiang Fan, Corentin Dumery, Nicolas Talabot, Pascal Fua
CrossDepth introduces geometry-constrained attention for multi-view surround depth estimation, addressing cross-image inconsistencies caused by varying camera intrinsics and limited receptive fields. The method conditions features on per-pixel camera-aware ray embeddings and extends pixel context via cross-image attention limited to geometrically plausible regions. Trained self-supervised with photometric consistency, it achieves better depth accuracy and consistency on DDAD and nuScenes compared to existing self-supervised approaches.
By Samer Abualhanud, Max Mehltretter
ARC‑Loc introduces a new cross‑view localization method that bypasses heavy Bird’s‑Eye‑View transformations and external depth models. By converting ground keypoints into azimuthal rays on a satellite map and exploiting their convergence at the user’s location, the approach uses a minimal Azimuthal Ray Convergence solver and an ARC loss to directly match ground and satellite images. Experiments on VIGOR and KITTI show that ARC‑Loc achieves competitive accuracy while offering faster, memory‑efficient inference and easy integration with existing frameworks.
By Hyeongsik Kim, Mincheol Kim, Heejoon Moon, Je Hyeong Hong