AI Paper Atlas

Research map / Benchmarks / LJSpeech

LJSpeech leaderboard

15 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on LJSpeech. Best MOS: 4.48 (BDDM - 12 steps, ICLR 2022).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

LJSpeech · MOS

15 papers, each with its best reported number for its own method; higher is better.

#MOS ↑MethodPaperTopic
14.48BDDM - 12 stepsBDDM: Bilateral Denoising Diffusion Models for Fast and High-Quality Speech Synthesis
ICLR 2022 · Table 1
Fast diffusion sampling
24.43VITSConditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech
ICML 2021 · Table 1
Variational and generative sampling
34.42DFlowDFlow: A Generative Model Combining Denoising AutoEncoder and Normalizing Flow for High Fidelity Waveform Generation
ICML 2024 · Table 2
Fast diffusion sampling
44.37EFTS-CNN (Mel+HiFi-GAN)EfficientTTS: An Efficient and High-Quality Text-to-Speech Architecture
ICML 2021 · MOS table, LJ-Speech
RNNs and normalization
54.36HiFi-GAN V1HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis
NeurIPS 2020 · Table: Comparison of the MOS a
GANs
64.258Avocodo V1
HiFi-GAN V1
Avocodo: Generative Adversarial Network for Artifact-Free Vocoder
AAAI 2023 · MOS results table
GANs
74.25PriorGrad, T_infer=50, 1M
DiffWave
PriorGrad: Improving Conditional Denoising Diffusion Models with Data-Dependent Adaptive Prior
ICLR 2022 · 5-scale subjective MOS table
Fast diffusion sampling
84.19RWEN-BERT
FastPitch w/ UAL
RWEN-TTS: Relation-Aware Word Encoding Network for Natural Text-to-Speech Synthesis
AAAI 2023 · Table: MOS evaluation results
Human and embodied modeling
94Cauchy Diffusion (η = 1)Cauchy Diffusion: A Heavy-tailed Denoising Diffusion Probabilistic Model for Speech Synthesis
AAAI 2025 · Table 2
Fast diffusion sampling
103.92DiffSpeechDiffSinger: Singing Voice Synthesis via Shallow Diffusion Mechanism
AAAI 2022 · TTS MOS table
Autoregressive and latent generation
5 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.