Readable Yet Unpredictable: Rotated-Outcome Prediction in Vision-Language Models

• URL: http://arxiv.org/abs/2606.07641v1
• Date: Mon, 01 Jun 2026 14:16:09 GMT
• Title: Readable Yet Unpredictable: Rotated-Outcome Prediction in Vision-Language Models
• Abstract summary: Can vision-language models predict what a 180 rotation would reveal from the original image alone?<n>We study this ability through Rotated-Outcome Prediction.<n>Current vision-language models can recognize a transformed visual state when it is shown, but often fail to predict that state from the original view.
• Score: 82.10380665213867
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Can vision-language models predict what a 180° rotation would reveal from the original image alone? We study this ability through Rotated-Outcome Prediction: given an original image, a model must answer what would be seen or read after a 180° in-plane rotation, without directly observing the rotated target. To isolate this gap, we introduce RotOutBench, a paired diagnostic benchmark spanning open visual cases and controlled text-image rotations. A sharp pattern emerges: many VLMs can recognize the relevant content when directly given either the original or rotated image, yet fail to infer the rotated result from the original image alone. On controlled text-image rotations, predicted-rotation accuracy collapses to near zero even for models with high direct-reading accuracy. A model-level case study further shows that the prediction state can approach a rotated-image reading state, while the final readout still shifts toward the original string. Current VLMs can recognize a transformed visual state when it is shown, but often fail to predict that state from the original view.

Related papers

• Are Image Generators Zero-Shot Perceivers? A Rigorous Evaluation [54.18771863630389]
  We ask how far image generators can go on public visual perception benchmarks in a zero-shot setting.<n>We introduce ProbeGen, a benchmark for zero-shot generative perception that casts monocular depth estimation.<n>We observe that pretrained image generators show measurable zero-shot perceptual competence, but with a clear trade-off.
  arXiv  Detail & Related papers  (2026-09-07T18:48:57Z)
• Seeing the Evidence, Missing the Answer: Tool-Guided Vision-Language Models on Visual Illusions [2.9521041993295607]
  Vision-language models (VLMs) exhibit a systematic bias when confronted with classic optical illusions.<n>We present a tool-guided inference framework that addresses this failure mode without any model training.<n>We report three empirical observations that we believe warrant additional investigation.
  arXiv  Detail & Related papers  (2026-03-31T08:35:10Z)
• RotBench: Evaluating Multimodal Large Language Models on Identifying Image Rotation [59.830657530592255]
  Multimodal Large Language Models (MLLMs) can accurately identify the orientation of input images rotated 0deg, 90deg, 180deg, and 270deg.<n>This task demands robust visual reasoning capabilities to detect rotational cues and contextualize spatial relationships within images, regardless of their orientation.<n>We show that several state-of-the-art open and proprietary MLLMs, including GPT-5, o3, and Gemini-2.5-Pro, do not reliably identify rotation in input images.
  arXiv  Detail & Related papers  (2025-08-19T15:58:25Z)
• RSAR: Restricted State Angle Resolver and Rotated SAR Benchmark [61.987291551925516]
  We introduce the Unit Cycle Resolver, which incorporates a unit circle constraint loss to improve angle prediction accuracy.<n>Our approach can effectively improve the performance of existing state-of-the-art weakly supervised methods.<n>With the aid of UCR, we further annotate and introduce RSAR, the largest multi-class rotated SAR object detection dataset to date.
  arXiv  Detail & Related papers  (2025-01-08T11:41:47Z)
• Breaking the Frame: Visual Place Recognition by Overlap Prediction [53.17564423756082]
  We propose a novel visual place recognition approach based on overlap prediction, called VOP.<n>VOP proceeds co-visible image sections by obtaining patch-level embeddings using a Vision Transformer backbone.<n>Our approach uses a voting mechanism to assess overlap scores for potential database images.
  arXiv  Detail & Related papers  (2024-06-23T20:00:20Z)
• Deep Rotation Correction without Angle Prior [57.76737888499145]
  We propose a new and practical task, named Rotation Correction, to automatically correct the tilt with high content fidelity. This task can be easily integrated into image editing applications, allowing users to correct the rotated images without any manual operations. We leverage a neural network to predict the optical flows that can warp the tilted images to be perceptually horizontal.
  arXiv  Detail & Related papers  (2022-07-07T02:46:27Z)
• Learning Continuous Rotation Canonicalization with Radial Beam Sampling [2.8935588665357077]
  We present a radial beam-based image canonicalization model, short BIC. Our model allows for maximal continuous angle regression and canonicalizes arbitrary center-rotated input images. As a pre-processing model, this enables rotation-invariant vision pipelines with model-agnostic rotation-sensitive downstream predictions.
  arXiv  Detail & Related papers  (2022-06-21T19:12:06Z)
• Outdoor inverse rendering from a single image using multiview self-supervision [36.065349509851245]
  We show how to perform scene-level inverse rendering to recover shape, reflectance and lighting from a single, uncontrolled image. The network takes an RGB image as input, regresses albedo, shadow and normal maps from which we infer least squares optimal spherical harmonic lighting. We believe this is the first attempt to use MVS supervision for learning inverse rendering.
  arXiv  Detail & Related papers  (2021-02-12T16:01:18Z)

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.