Sawtooth Sampling for Time Series Denoising Diffusion Implicit Models
This addresses the problem of slow sampling in diffusion models for time series generation, which is incremental as it builds on existing implicit diffusion models.
The paper tackles the computational expense of sampling in Denoising Diffusion Probabilistic Models (DDPMs) for time series data by introducing a Sawtooth Sampler that accelerates the reverse process, achieving a 30 times speed-up over the baseline while improving generated sequence quality for classification tasks.
Denoising Diffusion Probabilistic Models (DDPMs) can generate synthetic timeseries data to help improve the performance of a classifier, but their sampling process is computationally expensive. We address this by combining implicit diffusion models with a novel Sawtooth Sampler that accelerates the reverse process and can be applied to any pretrained diffusion model. Our approach achieves a 30 times speed-up over the standard baseline while also enhancing the quality of the generated sequences for classification tasks.