Beyond Cats and Dogs: Semi-supervised Classification of fuzzy labels with overclustering

Abstract

A long-standing issue with deep learning is the need for large and consistently labeled datasets. Although the current research in semi-supervised learning can decrease the required amount of annotated data by a factor of 10 or even more, this line of research still uses distinct classes like cats and dogs. However, in the real-world we often encounter problems where different experts have different opinions, thus producing fuzzy labels. We propose a novel framework for handling semi-supervised classifications of such fuzzy labels. Our framework is based on the idea of overclustering to detect substructures in these fuzzy labels. We propose a novel loss to improve the overclustering capability of our framework and show on the common image classification dataset STL-10 that it is faster and has better overclustering performance than previous work. On a real-world plankton dataset, we illustrate the benefit of overclustering for fuzzy labels and show that we beat previous state-of-the-art semisupervised methods. Moreover, we acquire 5 to 10% more consistent predictions of substructures. 1 * Corresponding Author 1 The source code, the datasets and supplementary material will be released soon. Please contact the corresponding author for any questions.

A Contrastive Evaluation Method for Discretion in Administrative Penalty

“…Schmarje L. et al [23] proposed a semi-supervised classification framework for processing fuzzy labels based on the idea of over-clustering. In this paper, a new loss algorithm using inverse cross-entropy was proposed to maximize mutual information.…”

Learning from fuzzy labels: Theoretical issues and algorithmic solutions

“…Despite the importance and practical relevance of learning from imprecise data in a variety of settings, so far research has mainly focused on specific tasks (in particular, semi-supervised learning and superset learning), while research on more general representations has been more limited. Furthermore, while several general-purpose algorithmic techniques (including generalized risk minimization [22,30,43,46,49,72], instance-based methods [4,7,51,79,82] or pseudo label-based learning [52,58,80]) have been developed to address these learning tasks, their theoretical and empirical properties have not yet been widely studied [11,15,56,59].…”

Credal Learning: Weakly Supervised Learning from Credal Sets

“…They have attracted interest in both the theoretical and application-oriented literature due to their flexibility as well as for the rich connections with convex analysis [12], optimization [2,35] and statistics [37]. Also in the context of machine learning (ML), credal sets and related models have recently attracted interest as a way to model weak supervision information in a variety of learning settings, including self-supervised learning [22,21], learning from noisy data [23,24], and learning from imprecise data [11,15,22], a general family of settings that encompasses, among others, semi-supervised learning, superset learning [17,26] and fuzzy label learning [10,32]. In all of these settings the idea is to model the weak supervision by means of credal sets, that are assumed to represent the partial or noisy information available to the annotating agent that produced the data: this general framework for studying weakly supervised learning is called credal learning.…”