BEV3SR is a training‑free, backbone‑agnostic framework that converts RGB‑D video observations into a queryable metric bird’s‑eye‑view (BEV) scene memory, enabling evidence‑aware 3D spatial reasoning. It formalizes each query into a spatial task specifying reference frame, constraints, objective, and evidence type, then routes the query to either a deterministic Geometric Reasoner or a Visual Evidence Aligner. Across multiple 3D scene QA benchmarks, BEV3SR consistently improves performance on spatially intensive and open‑ended questions within a 16‑frame budget.
arXiv:2506.17545v2 Announce Type: replace Abstract: Large language models are increasingly being applied to 3D scene understanding. However, existing 3D-aware LLMs largely operate as black boxes: the...
arXiv:2606. 11683v1 Announce Type: cross Abstract: Spatial reasoning from egocentric videos is inherently challenging because the observable evidence is constrained by the camera trajectory.
arXiv:2607. 02921v1 Announce Type: cross Abstract: Quantitative 3D spatial reasoning from egocentric RGB-D video is a critical capability for next-generation wearable assistants.
The paper introduces Vantage, a training‑free, model‑agnostic framework that pairs a vision‑language model with a 3D foundation model to perform Seek‑and‑View reasoning for multi‑view spatial understanding. Vantage first analyzes a question to plan a relevant viewpoint, then augments reasoning with geometry‑grounded visual evidence from that viewpoint. Experiments across six VLMs on five benchmarks show consistent improvements without fine‑tuning, demonstrating that revealing spatial evidence through view‑grounded reasoning reduces reliance on language‑based cross‑view alignment.
arXiv:2603. 04976v2 Announce Type: replace-cross Abstract: Reinforcement Learning with Verifiable Rewards ( RLVR ) has emerged as a transformative paradigm for enhancing the reasoning capabilities of Large Language Models ( LLMs), yet its potential in 3D scene understanding remains under-explored.
arXiv:2610.33558v1 Announce Type: new Abstract: Video spatial reasoning requires multimodal large language models (MLLMs) to understand how objects are arranged and related across viewpoints. While t...
arXiv:2607. 21595v1 Announce Type: cross Abstract: Despite rapid progress, most existing vision-language models (VLMs) built from 2D visual inputs often struggle when handling various 3D tasks that require fine-grained spatial understanding and reasoning.
arXiv:2609.36219v1 Announce Type: new Abstract: Perspective taking is a fundamental component of spatial intelligence, requiring models interpret spatial relations from a specified viewpoint, such as...
The paper introduces PanoVideoBench, a manually annotated benchmark with 1,000 questions covering five categories and 15 question types for panoramic video understanding, and identifies cross‑view tracking and sphere‑time evidence integration as weak points in existing models. To address these gaps, the authors create PanoVideoCoT, a dataset of 13,465 QA‑CoT pairs generated via sphere‑time‑grounded synthesis, and train PanoVideoReasoner using panoramic CoT fine‑tuning and evidence‑privileged on‑policy self‑distillation. The resulting model achieves 55.7% accuracy on PanoVideoBench, improving the baseline by 12.5 points and surpassing the previous state‑of‑the‑art by 2.6 points, particularly in spatial, cross‑view, and temporal reasoning.
OmniAct3D is a framework that adapts Vision Foundation Model (VFM) detectors trained on perspective images to equirectangular projection (ERP) panoramic images for 3D detection. It introduces the ERP-Ray Geometry Adapter (ERGA-Ray) to model spherical rays, the Visual-Action Reasoning Chain (VARC) to ground hypotheses in panoramic evidence, and the Appearance-Guided Heading Expert (AGHE) to recover high‑resolution heading cues. The method achieves a 2.96 NDS improvement on Spheriverse and a 24.87 mAP gain on PanoMMOcc, while preserving most of the baseline’s performance across configurations.
arXiv:2601. 19099v2 Announce Type: replace-cross Abstract: Vision--language models (VLMs) achieve strong performance on many multimodal benchmarks but remain brittle on spatial reasoning tasks that require aligning abstract overhead representations with egocentric views.
arXiv:2609.16233v1 Announce Type: cross Abstract: Vision-language models excel at 2D image understanding but remain limited in 3D spatial reasoning. Progress is hindered by limitations in current ben...