For on-device automatic speech recognition (ASR), quantization aware tra...
We present a novel sub-8-bit quantization-aware training (S8BQAT) scheme...
This work presents a scalable and efficient neural waveform codec (NWC) ...
In order to achieve high accuracy for machine learning (ML) applications...
Conventional audio coding technologies commonly leverage human perceptio...
Scalability and efficiency are desired in neural speech codecs, which
su...
In speech enhancement, an end-to-end deep neural network converts a nois...
In speech enhancement, an end-to-end deep neural network converts a nois...
Speech codecs learn compact representations of speech signals to facilit...
We present a psychoacoustically enhanced cost function to balance networ...
Given the progress in image recognition with recent data driven paradigm...