The reviewed record of science sign in
Pith

arxiv: 2203.15643 · v2 · pith:HVFDCXFP · submitted 2022-03-29 · cs.SD · cs.CL· cs.LG· cs.NE· eess.AS

Nix-TTS: Lightweight and End-to-End Text-to-Speech via Module-wise Distillation

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:HVFDCXFPrecord.jsonopen to challenge →

classification cs.SD cs.CLcs.LGcs.NEeess.AS
keywords distillationnix-ttsteacherend-to-endlightweightmodelmodule-wisenon-autoregressive
0
0 comments X
read the original abstract

Several solutions for lightweight TTS have shown promising results. Still, they either rely on a hand-crafted design that reaches non-optimum size or use a neural architecture search but often suffer training costs. We present Nix-TTS, a lightweight TTS achieved via knowledge distillation to a high-quality yet large-sized, non-autoregressive, and end-to-end (vocoder-free) TTS teacher model. Specifically, we offer module-wise distillation, enabling flexible and independent distillation to the encoder and decoder module. The resulting Nix-TTS inherited the advantageous properties of being non-autoregressive and end-to-end from the teacher, yet significantly smaller in size, with only 5.23M parameters or up to 89.34% reduction of the teacher model; it also achieves over 3.04x and 8.36x inference speedup on Intel-i7 CPU and Raspberry Pi 3B respectively and still retains a fair voice naturalness and intelligibility compared to the teacher model. We provide pretrained models and audio samples of Nix-TTS.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.