This repository contains the source code of our work on designing efficient CNNs for computer vision

Ukrainian TTS (text-to-speech) using ESPNET

Semantic segmentation on iPhone using ESPNetv2

EC499: Major Project

Pronunciation correction in vector quantized PPG representation space

This speech separation based framework is for multi-talker keyword spotting tasks and is implemented in the ESPnet2 toolkit.

Report of an end-to-end speech recognition task