{"ID":2886150,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2508.03118","arxiv_id":"2508.03118","title":"H3R: Hybrid Multi-view Correspondence for Generalizable 3D Reconstruction","abstract":"Despite recent advances in feed-forward 3D Gaussian Splatting, generalizable 3D reconstruction remains challenging, particularly in multi-view correspondence modeling. Existing approaches face a fundamental trade-off: explicit methods achieve geometric precision but struggle with ambiguous regions, while implicit methods provide robustness but suffer from slow convergence. We present H3R, a hybrid framework that addresses this limitation by integrating volumetric latent fusion with attention-based feature aggregation. Our framework consists of two complementary components: an efficient latent volume that enforces geometric consistency through epipolar constraints, and a camera-aware Transformer that leverages Plücker coordinates for adaptive correspondence refinement. By integrating both paradigms, our approach enhances generalization while converging 2$\\times$ faster than existing methods. Furthermore, we show that spatial-aligned foundation models (e.g., SD-VAE) substantially outperform semantic-aligned models (e.g., DINOv2), resolving the mismatch between semantic representations and spatial reconstruction requirements. Our method supports variable-number and high-resolution input views while demonstrating robust cross-dataset generalization. Extensive experiments show that our method achieves state-of-the-art performance across multiple benchmarks, with significant PSNR improvements of 0.59 dB, 1.06 dB, and 0.22 dB on the RealEstate10K, ACID, and DTU datasets, respectively. Code is available at https://github.com/JiaHeng-DLUT/H3R.","short_abstract":"Despite recent advances in feed-forward 3D Gaussian Splatting, generalizable 3D reconstruction remains challenging, particularly in multi-view correspondence modeling. Existing approaches face a fundamental trade-off: explicit methods achieve geometric precision but struggle with ambiguous regions, while implicit metho...","url_abs":"https://arxiv.org/abs/2508.03118","url_pdf":"https://arxiv.org/pdf/2508.03118v1","authors":"[\"Heng Jia\",\"Linchao Zhu\",\"Na Zhao\"]","published":"2025-08-05T05:56:30Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Transformer\",\"Variational Autoencoder\"]","has_code":false,"code_links":[{"ID":611278,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2886150,"paper_url":"https://arxiv.org/abs/2508.03118","paper_title":"H3R: Hybrid Multi-view Correspondence for Generalizable 3D Reconstruction","repo_url":"https://github.com/JiaHeng-DLUT/H3R","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}