Neural Multi-Speaker Voice Cloning for Nepali in Low-Resource Settings

TL;DR

This paper introduces a few-shot voice cloning system for Nepali, enabling speech synthesis of specific speakers with minimal data, addressing the low-resource challenge in Nepali speech technology.

Contribution

It develops a novel multi-component system combining speaker encoder, Tacotron2, and WaveRNN for Nepali voice cloning in low-resource settings, with new dataset construction and validation methods.

Findings

Effective cloning of unseen Nepali speakers.

Feasibility demonstrated with minimal data.

System achieves natural-sounding speech synthesis.

Abstract

This research presents a few-shot voice cloning system for Nepali speakers, designed to synthesize speech in a specific speaker's voice from Devanagari text using minimal data. Voice cloning in Nepali remains largely unexplored due to its low-resource nature. To address this, we constructed separate datasets: untranscribed audio for training a speaker encoder and paired text-audio data for training a Tacotron2-based synthesizer. The speaker encoder, optimized with Generative End2End loss, generates embeddings that capture the speaker's vocal identity, validated through Uniform Manifold Approximation and Projection (UMAP) for dimension reduction visualizations. These embeddings are fused with Tacotron2's text embeddings to produce mel-spectrograms, which are then converted into audio using a WaveRNN vocoder. Audio data were collected from various sources, including self-recordings, and underwent thorough preprocessing for quality and alignment. Training was performed using mel and gate loss functions under multiple hyperparameter settings. The system effectively clones speaker characteristics even for unseen voices, demonstrating the feasibility of few-shot voice cloning for the Nepali language and establishing a foundation for personalized speech synthesis in low-resource scenarios.