VoiceDiT: Dual-Condition Diffusion Transformer for Environment-Aware Speech Synthesis

TL;DR

VoiceDiT is a novel multi-modal diffusion transformer that generates environment-aware speech and audio from text and visual prompts, addressing alignment challenges in noisy conditions with improved quality.

Contribution

The paper introduces VoiceDiT, a new multi-modal generative model with a dual-condition diffusion transformer and a large-scale dataset for environment-aware speech synthesis.

Findings

Outperforms previous models on real-world datasets

Achieves better audio quality and environmental sound alignment

Effectively integrates multi-modal prompts in speech synthesis

Abstract

We present VoiceDiT, a multi-modal generative model for producing environment-aware speech and audio from text and visual prompts. While aligning speech with text is crucial for intelligible speech, achieving this alignment in noisy conditions remains a significant and underexplored challenge in the field. To address this, we present a novel audio generation pipeline named VoiceDiT. This pipeline includes three key components: (1) the creation of a large-scale synthetic speech dataset for pre-training and a refined real-world speech dataset for fine-tuning, (2) the Dual-DiT, a model designed to efficiently preserve aligned speech information while accurately reflecting environmental conditions, and (3) a diffusion-based Image-to-Audio Translator that allows the model to bridge the gap between audio and image, facilitating the generation of environmental sound that aligns with the multi-modal prompts. Extensive experimental results demonstrate that VoiceDiT outperforms previous models on real-world datasets, showcasing significant improvements in both audio quality and modality integration.