Stella X. Yu : Papers / Google Scholar

Open Ad-hoc Categorization with Contextualized Feature Learning
Zilin Wang and Sangwoo Mo and Stella X. Yu and Sima Behpour and Liu Ren
IEEE Conference on Computer Vision and Pattern Recognition, Nashville, Tennessee, 11-15 June 2025
Paper

Abstract

Adaptive categorization of visual scenes is essential for AI agents to handle changing tasks. Unlike fixed common categories for plants or animals, ad-hoc categories, such as things to sell at a garage sale, are created dynamically to achieve specific tasks. We study open ad-hoc categorization, where the goal is to infer novel concepts and categorize images based on a given context, a small set of labeled exemplars, and some unlabeled data.

We have two key insights: 1) recognizing ad-hoc categories relies on the same perceptual processes as common categories; 2) novel concepts can be discovered semantically by expanding contextual cues or visually by clustering similar patterns. We propose OAK, a simple model that introduces a single learnable context token into CLIP, trained with CLIP's objective of aligning visual and textual features and GCD's objective of clustering similar images.

On Stanford and Clevr-4 datasets, OAK consistently achieves the state-of-art in accuracy and concept discovery across multiple categorizations, including 87.4\% novel accuracy on Stanford Mood, surpassing CLIP and GCD by over 50\%. Moreover, OAK generates interpretable saliency maps, focusing on hands for Action, faces for Mood, and backgrounds for Location, promoting transparency and trust while enabling accurate and flexible categorization.


Keywords
{open ad-hoc categorization, contextualized feature learning, image classification, open-set recognition},