Tri-Bench: Stress-Testing VLM Reliability on Spatial Reasoning under Camera Tilt and Object Interference

• URL: http://arxiv.org/abs/2512.08860v1
• Date: Tue, 09 Dec 2025 17:52:57 GMT
• Title: Tri-Bench: Stress-Testing VLM Reliability on Spatial Reasoning under Camera Tilt and Object Interference
• Abstract summary: Verifiable geometric reasoning is a critical component for trustworthy and controllable agentic AI.<n>We present Tri-Bench, a compact benchmark of planar triangle problems.<n>We evaluate four recent Vision-Language Models (VLMs) using a single, fixed prompt.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Verifiable geometric reasoning is a critical component for trustworthy and controllable agentic AI. Despite impressive capabilities, Vision-Language Models (VLMs) often fail under realistic scene changes. We present Tri-Bench, a compact benchmark of planar triangle problems that isolates relative geometric reasoning while stressing two deployment-critical factors: camera pose (planar vs. tilted) and scene context via object interference (10 everyday objects). To test verifiability and control, we evaluate four recent VLMs using a single, fixed prompt whose guardrail explicitly describes a surrounding square border, enabling correct answers via homography. We evaluate six simple tasks over binary and continuous targets, and observe that the overall accuracy with respect to 3D ground truth is modest, ~69% on average (best ~75%, worst ~64%). The same responses align even more closely with 2D projections in the image plane, where mean accuracy is ~72%. All four VLMs consistently fail, with accuracy falling to ~0%, on recognizing minority shape classes (equilateral, isosceles, right-angled triangles). Additionally, overall VLM accuracy degrades by ~4.1% under camera tilt. This demonstrates that models fail to correctly utilize the explicit frame-of-reference hint provided in the prompt and default to 2D image plane cues. Finally, we find that object interference has no significant effect on VLM accuracy.

Related papers

• RigidBench: Evaluating Rigid-Body Physics in Video Generation Models [16.53417077740312]
  We introduce a simulator-grounded benchmark that compares a generated continuation with a reference rollout from the same initial frame and motion description.<n>We evaluate eight models on the same 100 examples with ten measurements that keep these aspects separate.<n>Full fine-tuning reduces 3D trajectory error by about 20% with almost no change in SSIM.
  arXiv  Detail & Related papers  (2026-08-16T06:05:20Z)
• MindEdit-Bench: Benchmarking Object-Level Counterfactual Spatial Reasoning in VLMs from In-the-Wild Photos [20.94052676416818]
  MindEdit-Bench is a benchmark of six spatial reasoning tasks built from three-photo smartphone triplets of newly captured indoor scenes.<n>Each question provides 8-24 structured answer choices, enabling answer-letter-level diagnosis of spatial and fallback errors.<n>The benchmark covers 120 private indoor scenes not drawn from public datasets.
  arXiv  Detail & Related papers  (2026-07-01T06:19:54Z)
• Beyond 3D VQAs: Injecting 3D Spatial Priors into Vision-Language Models for Enhanced Geometric Reasoning [24.251128353444713]
  Vision-Language Models (VLMs) often struggle with robust 3D spatial reasoning.<n>We propose GASP (GeometricAware Spatial Priors), a framework that injects fundamental geometric priors directly into the LLM's transformer layers.<n>GASP employs a small correspondence head, applied as a deep supervision signal across all layers, and is trained with a dual objective leveraging ground-scale video scenes.
  arXiv  Detail & Related papers  (2026-05-28T17:00:52Z)
• VLM-RobustBench: A Comprehensive Benchmark for Robustness of Vision-Language Models [64.56065206447788]
  Vision-language models (VLMs) achieve strong performance on standard, high-quality datasets.<n>We present VLM-RobustBench, a benchmark spanning 49 augmentation types across noise, blur, weather, digital, and geometrics.<n>We show that low-severity spatial perturbations often degrade performance more than visually severe photometric corruptions.
  arXiv  Detail & Related papers  (2026-03-06T10:58:02Z)
• PanoEnv: Exploring 3D Spatial Intelligence in Panoramic Environments with Reinforcement Learning [5.308328605042682]
  360 panoramic images are increasingly used in virtual reality, autonomous driving, and robotics for holistic scene understanding.<n>Current Vision-Language Models (VLMs) struggle with 3D spatial reasoning on Equirectangular Projection (ERP) images due to geometric distortion and limited 3D supervision.<n>We introduce PanoEnv, a large-scale VQA benchmark built from synthetic 3D environments.<n>Our 7B model achieves new state-of-the-art performance, improving overall accuracy to 52.93% (+3.59%) and open-ended accuracy to 14.83% while maintaining structured-task performance.
  arXiv  Detail & Related papers  (2026-02-25T15:12:17Z)
• Lost in Space? Vision-Language Models Struggle with Relative Camera Pose Estimation [37.0801171856967]
  Vision-Language Models (VLMs) perform well in 2D perception and semantic reasoning compared to their limited understanding of 3D spatial structure.<n>We introduce VRRPI-Bench, a benchmark derived from unlabeled egocentric videos with verbalized annotations of relative camera motion.<n>We also propose VRRPI-Diag, a diagnostic benchmark that isolates individual motion degrees of freedom.
  arXiv  Detail & Related papers  (2026-01-29T19:01:03Z)
• PRaDA: Projective Radial Distortion Averaging [40.77624901787694]
  We tackle the problem of automatic calibration of radially distorted cameras in challenging conditions.<n>Our proposed method, Projective Radial Distortion Averaging, averages multiple distortion estimates in a fully projective framework.
  arXiv  Detail & Related papers  (2025-04-23T08:22:59Z)
• Any6D: Model-free 6D Pose Estimation of Novel Objects [76.30057578269668]
  We introduce Any6D, a model-free framework for 6D object pose estimation.<n>It requires only a single RGB-D anchor image to estimate both the 6D pose and size of unknown objects in novel scenes.<n>We evaluate our method on five challenging datasets.
  arXiv  Detail & Related papers  (2025-03-24T13:46:21Z)
• Active 6D Pose Estimation for Textureless Objects using Multi-View RGB Frames [10.859307261818362]
  Estimating the 6D pose of textureless objects from RBG images is an important problem in robotics.<n>We propose a comprehensive active perception framework for estimating the 6D poses of textureless objects using only RGB images.
  arXiv  Detail & Related papers  (2025-03-05T18:28:32Z)
• DVMNet++: Rethinking Relative Pose Estimation for Unseen Objects [59.51874686414509]
  Existing approaches typically predict 3D translation utilizing the ground-truth object bounding box and approximate 3D rotation with a large number of discrete hypotheses.<n>We present a Deep Voxel Matching Network (DVMNet++) that computes the relative object pose in a single pass.<n>Our approach delivers more accurate relative pose estimates for novel objects at a lower computational cost compared to state-of-the-art methods.
  arXiv  Detail & Related papers  (2024-03-20T15:41:32Z)
• Learning to Estimate 6DoF Pose from Limited Data: A Few-Shot, Generalizable Approach using RGB Images [60.0898989456276]
  We present a new framework named Cas6D for few-shot 6DoF pose estimation that is generalizable and uses only RGB images. To address the false positives of target object detection in the extreme few-shot setting, our framework utilizes a self-supervised pre-trained ViT to learn robust feature representations. Experimental results on the LINEMOD and GenMOP datasets demonstrate that Cas6D outperforms state-of-the-art methods by 9.2% and 3.8% accuracy (Proj-5) under the 32-shot setting.
  arXiv  Detail & Related papers  (2023-06-13T07:45:42Z)
• Bridging Stereo Geometry and BEV Representation with Reliable Mutual Interaction for Semantic Scene Completion [45.171150395915056]
  3D semantic scene completion (SSC) is an ill-posed perception task that requires inferring a dense 3D scene from limited observations. Previous camera-based methods struggle to predict accurate semantic scenes due to inherent geometric ambiguity and incomplete observations. We resort to stereo matching technique and bird's-eye-view (BEV) representation learning to address such issues in SSC.
  arXiv  Detail & Related papers  (2023-03-24T12:33:44Z)
• Monocular 3D Object Detection with Depth from Motion [74.29588921594853]
  We take advantage of camera ego-motion for accurate object depth estimation and detection. Our framework, named Depth from Motion (DfM), then uses the established geometry to lift 2D image features to the 3D space and detects 3D objects thereon. Our framework outperforms state-of-the-art methods by a large margin on the KITTI benchmark.
  arXiv  Detail & Related papers  (2022-07-26T15:48:46Z)
• SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation [98.83762558394345]
  SO-Pose is a framework for regressing all 6 degrees-of-freedom (6DoF) for the object pose in a cluttered environment from a single RGB image. We introduce a novel reasoning about self-occlusion, in order to establish a two-layer representation for 3D objects. Cross-layer consistencies that align correspondences, self-occlusion and 6D pose, we can further improve accuracy and robustness.
  arXiv  Detail & Related papers  (2021-08-18T19:49:29Z)
• Uncertainty-Aware Camera Pose Estimation from Points and Lines [101.03675842534415]
  Perspective-n-Point-and-Line (Pn$PL) aims at fast, accurate and robust camera localizations with respect to a 3D model from 2D-3D feature coordinates.
  arXiv  Detail & Related papers  (2021-07-08T15:19:36Z)

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.