LSRM: High-Fidelity Object-Centric Reconstruction via Scaled Context Windows

• URL: http://arxiv.org/abs/2604.05182v1
• Date: Mon, 06 Apr 2026 21:21:12 GMT
• Title: LSRM: High-Fidelity Object-Centric Reconstruction via Scaled Context Windows
• Abstract summary: We introduce the Large Sparse Reconstruction Model to study how scaling transformer context windows impacts feed-forward 3D reconstruction.<n>We show that expanding the context window -- by substantially increasing the number of active object and image tokens -- remarkably narrows this gap and enables high-fidelity 3D object reconstruction and inverse rendering.
• Score: 10.300202521638274
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We introduce the Large Sparse Reconstruction Model to study how scaling transformer context windows impacts feed-forward 3D reconstruction. Although recent object-centric feed-forward methods deliver robust, high-quality reconstruction, they still lag behind dense-view optimization in recovering fine-grained texture and appearance. We show that expanding the context window -- by substantially increasing the number of active object and image tokens -- remarkably narrows this gap and enables high-fidelity 3D object reconstruction and inverse rendering. To scale effectively, we adapt native sparse attention in our architecture design, unlocking its capacity for 3D reconstruction with three key contributions: (1) an efficient coarse-to-fine pipeline that focuses computation on informative regions by predicting sparse high-resolution residuals; (2) a 3D-aware spatial routing mechanism that establishes accurate 2D-3D correspondences using explicit geometric distances rather than standard attention scores; and (3) a custom block-aware sequence parallelism strategy utilizing an All-gather-KV protocol to balance dynamic, sparse workloads across GPUs. As a result, LSRM handles 20x more object tokens and >2x more image tokens than prior state-of-the-art (SOTA) methods. Extensive evaluations on standard novel-view synthesis benchmarks show substantial gains over the current SOTA, yielding 2.5 dB higher PSNR and 40% lower LPIPS. Furthermore, when extending LSRM to inverse rendering tasks, qualitative and quantitative evaluations on widely-used benchmarks demonstrate consistent improvements in texture and geometry details, achieving an LPIPS that matches or exceeds that of SOTA dense-view optimization methods. Code and model will be released on our project page.

Related papers

• Lift3D-VLA: Lifting VLA Models to 3D Geometry and Dynamics-Aware Manipulation [76.16065370700488]
  We introduce Lift3D-VLA, a unified VLA framework that equips models with explicit 3D point cloud reasoning and enables temporally coherent action generation.<n>We show that Lift3D-VLA achieves 10.8% and 11.1% higher mean success rates on MetaWorld and RLBench than the best-performing prior VLA methods.
  arXiv  Detail & Related papers  (2026-07-07T17:59:47Z)
• Lite3R: A Model-Agnostic Framework for Efficient Feed-Forward 3D Reconstruction [23.223291360701907]
  Lite3R is a model-agnostic teacher-student framework that replaces dense attention with Sparse Linear Attention.<n>Lite3R substantially reduces latency (1.7-2.0x) and memory usage (1.9-2.4x) while preserving competitive reconstruction quality overall.
  arXiv  Detail & Related papers  (2026-05-12T00:20:02Z)
• Scal3R: Scalable Test-Time Training for Large-Scale 3D Reconstruction [50.5449251266956]
  This paper addresses the task of large-scale 3D scene reconstruction from long video sequences.<n>Recent feed-forward reconstruction models have shown promising results by directly regressing 3D geometry from RGB images without explicit 3D priors or geometric constraints.<n>We propose a novel neural global context representation that efficiently compresses and retains long-range scene information.
  arXiv  Detail & Related papers  (2026-04-09T17:59:50Z)
• Rewis3d: Reconstruction Improves Weakly-Supervised Semantic Segmentation [58.37682525044409]
  Rewis3d is a framework that leverages recent advances in feed-forward 3D reconstruction to significantly improve weakly supervised semantic segmentation on 2D images.<n>We show that Rewis3d achieves state-of-the-art performance in sparse supervision, outperforming existing approaches by 2-7% without requiring additional labels or inference overhead.
  arXiv  Detail & Related papers  (2026-03-06T15:25:20Z)
• AMB3R: Accurate Feed-forward Metric-scale 3D Reconstruction with Backend [18.645700170943975]
  AMB3R is a feed-forward model for dense 3D reconstruction on a metric-scale.<n>We show that AMB3R can be seamlessly extended to uncalibrated visual odometry (online) or large-scale structure from motion.
  arXiv  Detail & Related papers  (2025-11-25T14:23:04Z)
• LATTICE: Democratize High-Fidelity 3D Generation at Scale [27.310104395842075]
  LATTICE is a new framework for high-fidelity 3D asset generation.<n> VoxSet is a semi-structured representation that compresses 3D assets into a compact set of latent vectors anchored to a coarse voxel grid.<n>Our method is simple at its core, but supports arbitrary resolution decoding, low-cost training, and flexible inference schemes.
  arXiv  Detail & Related papers  (2025-11-24T03:22:19Z)
• LARM: A Large Articulated-Object Reconstruction Model [29.66486888001511]
  LARM is a unified feedforward framework that reconstructs 3D articulated objects from sparse-view images.<n>LARM generates auxiliary outputs such as depth maps and part masks to facilitate explicit 3D mesh extraction and joint estimation.<n>Our pipeline eliminates the need for dense supervision and supports high-fidelity reconstruction across diverse object categories.
  arXiv  Detail & Related papers  (2025-11-14T18:55:27Z)
• GauSSmart: Enhanced 3D Reconstruction through 2D Foundation Models and Geometric Filtering [50.675710727721786]
  We propose GauSSmart, a hybrid method that bridges 2D foundational models and 3D Gaussian Splatting reconstruction.<n>Our approach integrates established 2D computer vision techniques, including convex filtering and semantic feature supervision.<n>We validate our approach across three datasets, where GauSSmart consistently outperforms existing Gaussian Splatting.
  arXiv  Detail & Related papers  (2025-10-16T03:38:26Z)
• GeoLRM: Geometry-Aware Large Reconstruction Model for High-Quality 3D Gaussian Generation [65.33726478659304]
  We introduce the Geometry-Aware Large Reconstruction Model (GeoLRM), an approach which can predict high-quality assets with 512k Gaussians and 21 input images in only 11 GB GPU memory. Previous works neglect the inherent sparsity of 3D structure and do not utilize explicit geometric relationships between 3D and 2D images. GeoLRM tackles these issues by incorporating a novel 3D-aware transformer structure that directly processes 3D points and uses deformable cross-attention mechanisms.
  arXiv  Detail & Related papers  (2024-06-21T17:49:31Z)
• GEOcc: Geometrically Enhanced 3D Occupancy Network with Implicit-Explicit Depth Fusion and Contextual Self-Supervision [49.839374549646884]
  This paper presents GEOcc, a Geometric-Enhanced Occupancy network tailored for vision-only surround-view perception.<n>Our approach achieves State-Of-The-Art performance on the Occ3D-nuScenes dataset with the least image resolution needed and the most weightless image backbone.
  arXiv  Detail & Related papers  (2024-05-17T07:31:20Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.