[NeurIPS 2025] PyTorch implementation of [ThinkSound], a unified framework for generating audio from any modality, guided by Chain-of-Thought (CoT) reasoning.

Step-Audio 2 is an end-to-end multi-modal large language model designed for industry-strength audio understanding and speech conversation.

AcadHomepage: A Modern and Responsive Academic Personal Homepage

Kimi-Audio, an open-source audio foundation model excelling in audio understanding, generation, and conversation

Speech To Speech: an effort for an open-sourced and modular GPT4-o

VoiceBench: Benchmarking LLM-Based Voice Assistants

Multilingual Voice Understanding Model

Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.

VoiceLDM: Text-to-Speech with Environmental Context

jiazhenqi

Toolkits for Multimodal Emotion Recognition

official code for CVPR'24 paper Diff-BGM

Code for Novel View Acoustic Synthesis paper

DiffSinger: Singing Voice Synthesis via Shallow Diffusion Mechanism (SVS & TTS); AAAI 2022; Official code

PyTorch Implementation of ProDiff (ACM-MM'22) with a Extremely-Fast diffusion speech synthesis pipeline

A Non-Autoregressive Text-to-Speech (NAR-TTS) framework, including official PyTorch implementation of PortaSpeech (NeurIPS 2021) and DiffSpeech (AAAI 2022)

Repo for Visual Acoustic Matching, CVPR 2022

MUSIC-AVQA, CVPR2022 (ORAL)