SoLar: Sinkhorn Label Refinery for Imbalanced Partial-Label Learning
This addresses a combinatorial challenge in weakly-supervised learning for real-world applications with class imbalance, representing an incremental improvement over existing PLL methods.
The paper tackles the problem of partial-label learning (PLL) in imbalanced, long-tailed scenarios, where prior methods fail, and proposes SoLar, an Optimal Transport-based framework that refines disambiguated labels to match class priors, achieving substantially superior results on standardized benchmarks.
Partial-label learning (PLL) is a peculiar weakly-supervised learning task where the training samples are generally associated with a set of candidate labels instead of single ground truth. While a variety of label disambiguation methods have been proposed in this domain, they normally assume a class-balanced scenario that may not hold in many real-world applications. Empirically, we observe degenerated performance of the prior methods when facing the combinatorial challenge from the long-tailed distribution and partial-labeling. In this work, we first identify the major reasons that the prior work failed. We subsequently propose SoLar, a novel Optimal Transport-based framework that allows to refine the disambiguated labels towards matching the marginal class prior distribution. SoLar additionally incorporates a new and systematic mechanism for estimating the long-tailed class prior distribution under the PLL setup. Through extensive experiments, SoLar exhibits substantially superior results on standardized benchmarks compared to the previous state-of-the-art PLL methods. Code and data are available at: https://github.com/hbzju/SoLar .