The paper introduces SightCAD, a framework for parametric CAD reconstruction that explicitly handles view feasibility constraints in realistic indoor scenes. It learns a view selector to choose a small set of feasible viewpoints for a vision‑language model that generates executable CadQuery code, and jointly trains the selector and the VLM using a reward that depends on the final geometric fidelity. The approach outperforms baseline methods on several benchmarks, achieving higher intersection‑over‑union scores and a lower invalid‑code rate.
By Kartik Bali, Mahish Guru, Yiderigun Borjigin, Alexandra Starostina, Christian J. Cyron, Roland Aydin
PixelDense introduces a dual‑stream representation alignment for pixel diffusion, separating semantic and geometric teachers (DINOv2, SAM2, Depth Anything v2, Metric3D v2) into distinct projection spaces with an orthogonality penalty. The method improves dense‑prediction benchmarks, boosting PixelGen‑XXL’s GenEval score from 0.7927 to 0.8093, achieving significant gains in panoptic quality and depth accuracy, and accelerating training from random initialization. It also enhances SDEdit editing by preserving background structure and increasing PSNR.
By Lehan Yang, Daiqing Qi, Wenhao Zhang, Avery Li, Yiqing Yang, Yifan Li, Yu Kong, Haitian Zheng, Zhifei Zhang, Zhe Lin, Varun Jampani, Sheng Li
arXiv:2609.36374v1 Announce Type: new
Abstract: Feed-forward 3D reconstruction models have achieved impressive performance by scaling model and dataset size, but their cost excludes most research gro...
By Brandon Leblanc, Charalambos Poullis
The paper introduces a CAD‑free 3D shape prior that enhances object recognition by reconstructing each object with 3D Gaussian Splatting (3DGS) from short RGB‑D scans and fusing the resulting shape prototype with frozen DINOv2 image features. Experiments on T‑LESS and HOPE datasets show that geometry alone can match or exceed CAD‑based recognition, and that the combined approach improves performance, especially on shape‑distinctive or partially occluded objects. The study demonstrates that the benefit comes from the geometric information rather than rendered pixels, and that the prior is complementary to frozen vision features.
By Chenxi Tao, Seung-Kyum Choi
arXiv:2605. 10873v2 Announce Type: replace-cross Abstract: Recovering editable CAD programs from images or 3D observations is central to AI-assisted design, but progress is difficult to measure because existing evaluations are fragmented across datasets, modalities, and metrics.
By Anna C. Doris, Jacob Thomas Sony, Ghadi Nehme, Era Syla, Amin Heyrani Nobari, Faez Ahmed
Efficient Confidence-Aware Densification for 3DGS proposes a new densification strategy for 3D Gaussian Splatting that replaces opacity-based sampling with error-driven placement of primitives. The method keeps a fixed primitive budget and places new primitives where image error and low learned confidence are highest, relocating under-contributing primitives accordingly. Across 24 scenes, it outperforms existing fixed-budget methods—including 3DGS-MCMC, ConeGS, and EDGS—in PSNR, SSIM, and LPIPS at budgets from 100k to 2M primitives, and can also leverage prior-based initialization to match or exceed prior methods.
By Manuel Hofer, Thomas K\"ohler, Felix Windisch, Bernhard Kerbl, Markus Steinberger, Lukas Radl
arXiv:2609.01306v1 Announce Type: cross
Abstract: Triangle-based neural rendering bridges neural scene representations and conventional graphics pipelines by optimizing explicit geometric primitives...
By Kaixuan Zhang, Minxian Li, Mingwu Ren, Xiatian Zhu
arXiv:2607. 09825v1 Announce Type: cross Abstract: Robotic manipulation policies rely on pre-trained vision models that give either a global scene embedding or a dense patch grid.
By Yi Li (TU Darmstadt), Alexandre Chapin (LIRIS), Liming Chen (LIRIS), Jan Peters (TU Darmstadt), Alap Kshirsagar (IIT Delhi, ADU)
Groundbench is a new benchmark that evaluates vision‑language models on multi‑resolution polygon grounding, using the same 1,500 image‑expression‑referent triples but targeting exact‑N polygons with five different vertex budgets. It audits both filled‑region intersection‑over‑union (IoU) and legal‑polygon completion, revealing that the best models achieve 88.2 box IoU and 97.1 accuracy at IoU ≥ 0.5, while direct polygon predictions lag at 57.7 and 69.2. The study shows performance is non‑monotonic across budgets, collapses at the densest budget due to legality failures, and highlights that false spatial cues hurt more than false colour cues, underscoring an operational geometry gap beyond latent boundary perception.
By Zhonghan Bian, Zhenran Wang, Jinsong Li, Zhangyang Qi
arXiv:2608.21008v1 Announce Type: new
Abstract: Mobile AR frameworks attach a metric pose prior to every casual phone capture, and turning it into reconstruction-grade poses cheaply on CPU is the ste...
By Nikolaos Kyriazis
Budgeted-GS is a post‑hoc technique that transforms any trained 3D Gaussian Splatting model into a factoring tree—a multi‑resolution hierarchy of moment‑matched aggregates. After a brief construction pass, a single quality parameter selects the appropriate level of detail for each view, allowing the same city‑scale model to run in real time on GPUs with varying memory capacities. The method also informs budget‑centered training, enabling models to be built directly at the necessary size and avoiding wasted optimization of later discarded primitives.
By Haipeng Wang
arXiv:2605. 07066v3 Announce Type: replace Abstract: Autonomous systems that build structures from natural-language instructions need reliable spatial reasoning, yet large language models (LLMs) make systematic coordinate errors when generating three-dimensional block placements.
By Paul Whitten, Li-Jen Chen, Sharath Baddam