AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

StyleTTS 2: Towards Human-Level Text-to-Speech through Style Diffusion and Adversarial Training with Large Speech Language Models

Yinghao Aaron Li, Cong Han, Vinay S Raghavan, Gavin Mischler, Nima Mesgarani

NeurIPS 2023 · Image generation › Autoregressive and latent generation · approach: Neural text-to-speech synthesis · 24 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add style diffusion (vs random encoded styles) (ablation)StyleTTS 2 · LJSpeech OOD textsCMOS-N: -0.46Improved
add SLM (WavLM) adversarial training (ablation)StyleTTS 2 · LJSpeech OOD textsCMOS-N: -0.32Improved
add prosodic style encoder (ablation)StyleTTS 2 · LJSpeech OOD textsCMOS-N: -0.35Improved
add differentiable duration upsampler (ablation)StyleTTS 2 · LJSpeech OOD textsCMOS-N: -0.21Improved
add OOD texts to SLM adversarial training (ablation)StyleTTS 2 · LJSpeech OOD textsCMOS-N: -0.15Improved
add SLM adversarial training (ablation)StyleTTS 2 · LJSpeechWER: WER 6.52 -> 6.50No clear effect
add SLM adversarial training (ablation)StyleTTS 2 · LJSpeechDUR MAD: DUR MAD 0.513 (w/o) vs 0.521Hurt
add differentiable upsampler (ablation)StyleTTS 2 · LJSpeechF0 RMSE: F0 RMSE 0.880 -> 0.651Improved
add style diffusion (ablation)StyleTTS 2 · LJSpeechWER: WER 8.77% -> 6.50%Improved
increase style diffusion steps beyond 3-5 (ablation)StyleTTS 2 · LJSpeechMCDNo clear effect
use StyleTTS 2 instead of ground truth recordingsTTS system · LJSpeechCMOS-N: 0.28Improved
use StyleTTS 2 vs ground truthTTS system · VCTKCMOS-N: -0.02No clear effect

Later papers that compared with this method

2 later accepted papers report StyleTTS2 as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy