Learning from Ambiguously Labeled Face Images

TL;DR

This paper introduces a novel matrix completion-based method, WMCar, for resolving ambiguous labels in face images, and an iterative candidate elimination technique to improve label disambiguation and classification accuracy.

Contribution

The paper proposes WMCar and ICE methods that enhance label disambiguation in ambiguous face image datasets, incorporating label imbalance handling and prior constraints.

Findings

WMCar improves label disambiguation accuracy.

ICE effectively reduces ambiguity through iterative candidate elimination.

The approach outperforms existing methods on multiple datasets.

Abstract

Learning a classifier from ambiguously labeled face images is challenging since training images are not always explicitly-labeled. For instance, face images of two persons in a news photo are not explicitly labeled by their names in the caption. We propose a Matrix Completion for Ambiguity Resolution (MCar) method for predicting the actual labels from ambiguously labeled images. This step is followed by learning a standard supervised classifier from the disambiguated labels to classify new images. To prevent the majority labels from dominating the result of MCar, we generalize MCar to a weighted MCar (WMCar) that handles label imbalance. Since WMCar outputs a soft labeling vector of reduced ambiguity for each instance, we can iteratively refine it by feeding it as the input to WMCar. Nevertheless, such an iterative implementation can be affected by the noisy soft labeling vectors, and thus the performance may degrade. Our proposed Iterative Candidate Elimination (ICE) procedure makes the iterative ambiguity resolution possible by gradually eliminating a portion of least likely candidates in ambiguously labeled face. We further extend MCar to incorporate the labeling constraints between instances when such prior knowledge is available. Compared to existing methods, our approach demonstrates improvement on several ambiguously labeled datasets.