Dendrogram distance: an evaluation metric for generative networks using hierarchical clustering
This addresses the need for better evaluation metrics in generative modeling, particularly for detecting mode collapse, but it is incremental as it builds on existing clustering and divergence approaches.
The paper tackles the problem of evaluating generative networks by introducing a novel metric based on hierarchical clustering dendrograms to measure divergence between real and generated data, specifically targeting mode collapse, and shows it is competitive with state-of-the-art methods in controlled evaluations.
We present a novel metric for generative modeling evaluation, focusing primarily on generative networks. The method uses dendrograms to represent real and fake data, allowing for the divergence between training and generated samples to be computed. This metric focus on mode collapse, targeting generators that are not able to capture all modes in the training set. To evaluate the proposed method it is introduced a validation scheme based on sampling from real datasets, therefore the metric is evaluated in a controlled environment and proves to be competitive with other state-of-the-art approaches.