FFAD: A Novel Metric for Assessing Generated Time Series Data Utilizing Fourier Transform and Auto-encoder
This addresses a gap in assessment capabilities for generative time series data, which is incremental as it adapts existing concepts to a new domain.
The paper tackles the lack of a standard metric for evaluating generative models on time series data by proposing FFAD, a novel metric using Fourier transform and auto-encoder, which effectively distinguishes samples from different classes in experiments.
The success of deep learning-based generative models in producing realistic images, videos, and audios has led to a crucial consideration: how to effectively assess the quality of synthetic samples. While the Fréchet Inception Distance (FID) serves as the standard metric for evaluating generative models in image synthesis, a comparable metric for time series data is notably absent. This gap in assessment capabilities stems from the absence of a widely accepted feature vector extractor pre-trained on benchmark time series datasets. In addressing these challenges related to assessing the quality of time series, particularly in the context of Fréchet Distance, this work proposes a novel solution leveraging the Fourier transform and Auto-encoder, termed the Fréchet Fourier-transform Auto-encoder Distance (FFAD). Through our experimental results, we showcase the potential of FFAD for effectively distinguishing samples from different classes. This novel metric emerges as a fundamental tool for the evaluation of generative time series data, contributing to the ongoing efforts of enhancing assessment methodologies in the realm of deep learning-based generative models.