Controlling the Fidelity and Diversity of Deep Generative Models via Pseudo Density
This addresses the challenge of balancing fidelity and diversity in generative models for AI researchers and practitioners, offering incremental improvements through novel techniques.
The paper tackles the problem of controlling fidelity and diversity in deep generative models like GANs and diffusion models, introducing a pseudo density metric based on nearest-neighbor information to adjust these properties, and reports improved Frechet Inception Distance (FID) for pre-trained models with minimal fine-tuning iterations.
We introduce an approach to bias deep generative models, such as GANs and diffusion models, towards generating data with either enhanced fidelity or increased diversity. Our approach involves manipulating the distribution of training and generated data through a novel metric for individual samples, named pseudo density, which is based on the nearest-neighbor information from real samples. Our approach offers three distinct techniques to adjust the fidelity and diversity of deep generative models: 1) Per-sample perturbation, enabling precise adjustments for individual samples towards either more common or more unique characteristics; 2) Importance sampling during model inference to enhance either fidelity or diversity in the generated data; 3) Fine-tuning with importance sampling, which guides the generative model to learn an adjusted distribution, thus controlling fidelity and diversity. Furthermore, our fine-tuning method demonstrates the ability to improve the Frechet Inception Distance (FID) for pre-trained generative models with minimal iterations.