Human-Like Coarse Object Representations in Vision Models

2026-02-12

Technology Category

Computer Vision: Visual Reasoning & Symbolic RepresentationsKnowledge Representation and Reasoning: Common-Sense ReasoningMachine Learning: Representation Learning

Application Category

Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactionsEconomics, Online Markets and Human Computation: Cost models of using LLMs in production systemsUser Modeling, Personalization and Recommendation: Practical large-scale studies of user experience

Abstract

Humans appear to represent objects for intuitive physics with coarse, volumetric bodies''that smooth concavities - trading fine visual details for efficient physical predictions - yet their internal structure is largely unknown. Segmentation models, in contrast, optimize pixel-accurate masks that may misalign with such bodies. We ask whether and when these models nonetheless acquire human-like bodies. Using a time-to-collision (TTC) behavioral paradigm, we introduce a comparison pipeline and alignment metric, then vary model training time, size, and effective capacity via pruning. Across all manipulations, alignment with human behavior follows an inverse U-shaped curve: small/briefly trained/pruned models under-segment into blobs; large/fully trained models over-segment with boundary wiggles; and an intermediate ideal body granularity''best matches humans. This suggests human-like coarse bodies emerge from resource constraints rather than bespoke biases, and points to simple knobs - early checkpoints, modest architectures, light pruning - for eliciting physics-efficient representations. We situate these results within resource-rational accounts balancing recognition detail against physical affordances.

Problem

• coarse object representations
• intuitive physics
• human-like representations
• segmentation models
• resource constraints

Innovation

• coarse object representations
• intuitive physics
• resource-rationality
• model pruning
• time-to-collision