Stacked U-Net with High-level Feature Transfer for Parameter Efficient Speech Enhancement

LiteTTS: A Decoder-free Light-weight Text-to-wave Synthesis Based on Generative Adversarial Networks

Disentangled Representations for Arabic Dialect Identification based on Supervised Clustering with Triplet Loss

Self-supervised Complex Network for Machine Sound Anomaly Detection

A Fast and Lightweight Text-To-Speech Model with Spectrum and Waveform Alignment Algorithms