Describing objects by their attributes

TLDR

Learning attributes requires generalizing across object categories, a challenge beyond instance‑level variation. The study aims to shift object recognition from naming to attribute description, treating attribute inference as the core problem. They introduce a feature‑selection approach that learns attributes generalizable across categories, enabling description of familiar and unfamiliar objects with few or no examples. Evaluation demonstrates that the attribute‑based framework outperforms the traditional naming paradigm, revealing its limitations and highlighting new descriptive capabilities.

Abstract

We propose to shift the goal of recognition from naming to describing. Doing so allows us not only to name familiar objects, but also: to report unusual aspects of a familiar object ("spotty dog", not just "dog"); to say something about unfamiliar objects ("hairy and four-legged", not just "unknown"); and to learn how to recognize new objects with few or no visual examples. Rather than focusing on identity assignment, we make inferring attributes the core problem of recognition. These attributes can be semantic ("spotty") or discriminative ("dogs have it but sheep do not"). Learning attributes presents a major new challenge: generalization across object categories, not just across instances within a category. In this paper, we also introduce a novel feature selection method for learning attributes that generalize well across categories. We support our claims by thorough evaluation that provides insights into the limitations of the standard recognition paradigm of naming and demonstrates the new abilities provided by our attribute-based framework.