本页面归档日常积累的技术分享和 Talk Slides，涵盖语音合成、语音识别、音频生成等多个方向。
2026
21
VTP: Towards Scalable Pre-training of Visual Tokenizers for Generation
关键工作：VTP, Visual Tokenizer, RAE
Slides
20
LongCat-AudioDiT: High-Fidelity Diffusion Text-to-Speech in the Waveform Latent Space
关键工作：LongCat-AudioDiT, Wav-VAE, APG
Slides
2025
19
Recent Advances in Autoregressive Diffusion Models for Speech Generation
关键工作：MELA-TTS, VoxCPM, VibeVoice, Ming-UniAudio
Slides
18
DraftDiffRO: Differentiable Reward Optimization for LLM based TTS system
关键工作：CosyVoice, CosyVoice2
Slides
17
16
MegaTTS 3: Sparse Alignment Enhanced Latent Diffusion Transformer for Zero-Shot Speech Synthesis
关键工作：MegaTTS 3, InspireMusic, RALL-E, VALLE-R, VoiceLDM, PeRFlow
Slides
2024
15
FireRedTTS: A Foundation Text-To-Speech Framework for Industry-Level Generative Speech Applications
关键工作：FireRedTTS, Semantic-Aware Speech Tokenizer
Slides
14
13
CosyVoice: A Scalable Multilingual Zero-shot Text-to-speech Synthesizer based on Supervised Semantic Tokens
关键工作：CosyVoice
Slides
12
NaturalSpeech 3: Zero-Shot Speech Synthesis with Factorized Codec and Diffusion Models
关键工作：FACodec, NaturalSpeech 3, MaskGIT, Discrete Diffusion
Slides
11
MobileSpeech: A Fast and High-Fidelity Framework for Mobile Zero-Shot Text-to-Speech
关键工作：MobileSpeech, SoundStorm, Pheme
Slides
2023
10
Recent Advances on Unified Model for Voice and Audio Generation (Part 2)
关键工作：AudioLDM 2, Stable Diffusion
Slides
9
Recent Advances on Unified Model for Voice and Audio Generation (Part 1)
关键工作：Speech-X, Make-A-Voice / MVoice, UniAudio
Slides
8
Speak, Read and Prompt: High-Fidelity Text-to-Speech with Minimal Supervision
关键工作：w2v-BERT, SoundStream, Spear-TTS
Slides
7
Text-Based Speech Editing From a Perspective of Mask-Based Generative Models
关键工作：SpeechPainter, A3T, MUSE (文生图)
Slides
2022
6
5
Voice Cloning B: Overview on Transfer Learning, Representation Learning and Meta-Learning
关键工作：ProsodySpeech, StyleSpeech, GST, Meta-TTS
Slides
4
Voice Cloning A: Recent Advances in Adaptive Text-to-Speech
关键工作：AdaSpeech, AdaSpeech2, AdaSpeech3, AdaSpeech4
Slides
2021
3
2020
2
An Overview of Hybrid ASR: Acoustic Models, Language Models, and Post-Processing
关键工作：LF-MMI, TDNN-F, Lattice Rescoring
Slides
1
Personal Work on Speech Recognition (Bachelor/Master)
关键工作：Speaker Adaption, Transfer Learning
Slides