The Last Visible Pixel: Probing Fine-Scale Perception in Vision-Language Models
Recent vision-language models (VLMs) excel at multimodal understanding and reasoning, yet their fine-grained visual perception remains underexplored. A natural extension of ``How many r are there in Strawberry?'' asks: how small a visual pattern can a VLM reliably perceive? As such, we introduce FineSightBench, a new benchmark that systematically probes this limit by separating perception tasks (pixel-level recognition of letters, shapes, objects) from reasoning tasks (spatial reasoning, counting, ordering over small targets) across controlled scales of 4--48px. Through comprehensive experiments and detailed failure mode analysis on state-of-the-art models, we reveal a sharp dissociation: perception saturates around 12px, while reasoning remains limited even at larger scales, with persistent numeracy and sequence errors. These findings expose fundamental deficiencies in VLMs' fine-scale visual reasoning that demand more rigorous evaluation.
Code (0)
등록된 구현이 없습니다.
Tasks
Spatial ReasoningVisual ReasoningSimilar Papers 제목 키워드 기반
Everything that can be learned about a causal structure with latent variables by observational and interventional probing schemes
What types of differences among causal structures with latent variables are impossible to distinguish by statistical data obtained by probing each visible variable? If the probing scheme is simply passive observation, th…
XoFTR: Cross-modal Feature Matching Transformer
We introduce, XoFTR, a cross-modal cross-view method for local feature matching between thermal infrared (TIR) and visible images. Unlike visible images, TIR images are less susceptible to adverse lighting and weather co…
Image AugmentationAdaptive Assignment for Geometry Aware Local Feature Matching
The detector-free feature matching approaches are currently attracting great attention thanks to their excellent performance. However, these methods still struggle at large-scale and viewpoint variations, due to the geom…
Feature CorrelationGenerative adversarial network based single pixel imaging
Single pixel imaging can reconstruct two-dimensional images of a scene with only a single-pixel detector. It has been widely used for imaging in non-visible bandwidth (e.g., near-infrared and X-ray) where focal-plane arr…
Generative Adversarial NetworkA Synthesis-Based Approach for Thermal-to-Visible Face Verification
In recent years, visible-spectrum face verification systems have been shown to match the performance of experienced forensic examiners. However, such systems are ineffective in low-light and nighttime conditions. Thermal…
Face AlignmentFace GenerationFace Verification