{"work":{"id":"c5e225cf-22b3-4efa-a3c4-5346587616d0","openalex_id":"https://openalex.org/W4404261781","doi":"10.48550/arxiv.2410.17242","arxiv_id":"2410.17242","raw_key":null,"title":"LVSM: A Large View Synthesis Model with Minimal 3D Inductive Bias","authors":null,"authors_text":"Haian Jin, Hanwen Jiang, Hao Tan, Kai Zhang, Sai Bi, Tianyuan Zhang, Fujun Luan, Noah Snavely, and Zexiang Xu","year":2024,"venue":"cs.CV","abstract":"We propose the Large View Synthesis Model (LVSM), a novel transformer-based approach for scalable and generalizable novel view synthesis from sparse-view inputs. We introduce two architectures: (1) an encoder-decoder LVSM, which encodes input image tokens into a fixed number of 1D latent tokens, functioning as a fully learned scene representation, and decodes novel-view images from them; and (2) a decoder-only LVSM, which directly maps input images to novel-view outputs, completely eliminating intermediate scene representations. Both models bypass the 3D inductive biases used in previous methods -- from 3D representations (e.g., NeRF, 3DGS) to network designs (e.g., epipolar projections, plane sweeps) -- addressing novel view synthesis with a fully data-driven approach. While the encoder-decoder model offers faster inference due to its independent latent representation, the decoder-only LVSM achieves superior quality, scalability, and zero-shot generalization, outperforming previous state-of-the-art methods by 1.5 to 3.5 dB PSNR. Comprehensive evaluations across multiple datasets demonstrate that both LVSM variants achieve state-of-the-art novel view synthesis quality. Notably, our models surpass all previous methods even with reduced computational resources (1-2 GPUs). Please see our website for more details: https://haian-jin.github.io/projects/LVSM/ .","external_url":"https://arxiv.org/abs/2410.17242","cited_by_count":0,"metadata_source":"pith","metadata_fetched_at":"2026-08-05T02:28:24.338817+00:00","pith_arxiv_id":"2410.17242","created_at":"2026-05-10T10:55:03.916603+00:00","updated_at":"2026-08-05T02:28:24.338817+00:00","title_quality_ok":true,"display_title":"Lvsm: A large view synthesis model with minimal 3d inductive bias","render_title":"Lvsm: A large view synthesis model with minimal 3d inductive bias"},"hub":{"state":{"work_id":"c5e225cf-22b3-4efa-a3c4-5346587616d0","tier":"hub","tier_reason":"10+ Pith inbound or 1,000+ external citations","pith_inbound_count":23,"external_cited_by_count":0,"distinct_field_count":4,"first_pith_cited_at":"2025-05-29T17:50:34+00:00","last_pith_cited_at":"2026-07-08T09:02:18+00:00","author_build_status":"not_needed","summary_status":"needed","contexts_status":"needed","graph_status":"needed","ask_index_status":"not_needed","reader_status":"not_needed","recognition_status":"not_needed","updated_at":"2026-08-22T02:59:45.575132+00:00","tier_text":"hub"},"tier":"hub","role_counts":[{"context_role":"background","n":4},{"context_role":"baseline","n":2}],"polarity_counts":[{"context_polarity":"background","n":4},{"context_polarity":"baseline","n":2}],"runs":{},"summary":{},"graph":{},"authors":[]}}