GICDM: Mitigating Hubness for Reliable Distance-Based Generative Model Evaluation
This addresses a reliability issue in evaluating generative models for researchers and practitioners, but it is incremental as it builds on the classical ICDM method.
The paper tackles the problem of hubness distorting nearest neighbor relationships in high-dimensional embedding spaces used for generative model evaluation, and introduces GICDM to correct this, showing it resolves failures and improves alignment with human judgment in experiments.
Generative model evaluation commonly relies on high-dimensional embedding spaces to compute distances between samples. We show that dataset representations in these spaces are affected by the hubness phenomenon, which distorts nearest neighbor relationships and biases distance-based metrics. Building on the classical Iterative Contextual Dissimilarity Measure (ICDM), we introduce Generative ICDM (GICDM), a method to correct neighborhood estimation for both real and generated data. We introduce a multi-scale extension to improve empirical behavior. Extensive experiments on synthetic and real benchmarks demonstrate that GICDM resolves hubness-induced failures, restores reliable metric behavior, and improves alignment with human judgment.