Dual-level Fuzzy Learning with Patch Guidance for Image Ordinal Regression

2025-05-09

AI Summary

Existing image-level ordinal regression methods rely solely on coarse-grained ordinal labels, failing to capture fine-grained, discriminative patch-level features that human experts routinely exploit. To address this, we propose a weakly supervised fine-grained ordinal regression framework: it implicitly generates high-quality patch-level supervision signals—without explicit patch annotations—via implicit patch labeling and adaptive filtering. Furthermore, it introduces channel-patch dual-dimensional fuzzy modeling and a hierarchical loss design to achieve precise and robust ordinal boundary prediction. To our knowledge, this is the first method to enable fine-grained ordinal boundary learning exclusively from image-level labels. Extensive experiments on multiple benchmark datasets demonstrate substantial improvements over state-of-the-art methods, particularly in distinguishing adjacent ordinal classes—a notoriously challenging case. The source code is publicly available.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Object Detection & Categorization
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Ordinal regression bridges regression and classification by assigning objects to ordered classes. While human experts rely on discriminative patch-level features for decisions, current approaches are limited by the availability of only image-level ordinal labels, overlooking fine-grained patch-level characteristics. In this paper, we propose a Dual-level Fuzzy Learning with Patch Guidance framework, named DFPG that learns precise feature-based grading boundaries from ambiguous ordinal labels, with patch-level supervision. Specifically, we propose patch-labeling and filtering strategies to enable the model to focus on patch-level features exclusively with only image-level ordinal labels available. We further design a dual-level fuzzy learning module, which leverages fuzzy logic to quantitatively capture and handle label ambiguity from both patch-wise and channel-wise perspectives. Extensive experiments on various image ordinal regression datasets demonstrate the superiority of our proposed method, further confirming its ability in distinguishing samples from difficult-to-classify categories. The code is available at https://github.com/ZJUMAI/DFPG-ord.

Problem

• Bridges regression and classification with ordered classes
• Overlooks fine-grained patch-level features in ordinal labels
• Handles label ambiguity using dual-level fuzzy learning

Innovation

• Dual-level fuzzy learning with patch guidance
• Patch-labeling and filtering strategies for feature focus
• Fuzzy logic for handling label ambiguity quantitatively