DPI-TTS: Directional Patch Interaction for Fast-Converging and Style Temporal Modeling in Text-to-Speech

TL;DR

DPI-TTS introduces a novel speech diffusion model that accelerates training and enhances naturalness by incorporating directional patch interactions and style-aware temporal modeling tailored to speech's acoustic properties.

Contribution

It presents DPI-TTS, a new method that improves training speed and speech naturalness by integrating directional patch interactions and style temporal modeling in diffusion-based TTS.

Findings

Training speed increased by nearly 2 times

Significant improvement in speech naturalness

Enhanced speaker style similarity

Abstract

In recent years, speech diffusion models have advanced rapidly. Alongside the widely used U-Net architecture, transformer-based models such as the Diffusion Transformer (DiT) have also gained attention. However, current DiT speech models treat Mel spectrograms as general images, which overlooks the specific acoustic properties of speech. To address these limitations, we propose a method called Directional Patch Interaction for Text-to-Speech (DPI-TTS), which builds on DiT and achieves fast training without compromising accuracy. Notably, DPI-TTS employs a low-to-high frequency, frame-by-frame progressive inference approach that aligns more closely with acoustic properties, enhancing the naturalness of the generated speech. Additionally, we introduce a fine-grained style temporal modeling method that further improves speaker style similarity. Experimental results demonstrate that our method increases the training speed by nearly 2 times and significantly outperforms the baseline models.