An Improved Evaluation Framework for Generative Adversarial Networks
This work addresses the need for better evaluation metrics in GAN research, particularly for domain-specific applications, though it is incremental as it builds on existing methods like FID.
The paper tackles the problem of evaluating Generative Adversarial Networks (GANs) for domain-specific images by proposing an improved framework that uses a specialized encoder for fine-grained representation and a Class-Aware Fréchet Distance (CAFD) metric, showing it overcomes shortcomings of the FID method and improves robustness.
In this paper, we propose an improved quantitative evaluation framework for Generative Adversarial Networks (GANs) on generating domain-specific images, where we improve conventional evaluation methods on two levels: the feature representation and the evaluation metric. Unlike most existing evaluation frameworks which transfer the representation of ImageNet inception model to map images onto the feature space, our framework uses a specialized encoder to acquire fine-grained domain-specific representation. Moreover, for datasets with multiple classes, we propose Class-Aware Frechet Distance (CAFD), which employs a Gaussian mixture model on the feature space to better fit the multi-manifold feature distribution. Experiments and analysis on both the feature level and the image level were conducted to demonstrate improvements of our proposed framework over the recently proposed state-of-the-art FID method. To our best knowledge, we are the first to provide counter examples where FID gives inconsistent results with human judgments. It is shown in the experiments that our framework is able to overcome the shortness of FID and improves robustness. Code will be made available.