Phoneme recognition is a very important part of speech recognition that requires the ability
to extract phonetic features from multiple frames. In this paper, we compare and analyze …

Streaming automatic speech recognition (ASR) models are restricted from accessing future
context, which results in worse performance compared to the non-streaming models. To …

Pre-training of self-supervised learning (SSL) generally shows a good performance on
various speech processing tasks. However, this pre-training scheme may lead to a sub …

Self-training has gained increasing attention due to its notable performance improvement in
speech recognition. However, conventional self-training techniques have two key …

Analog Content Addressable Memories (aCAMs) have proven useful for associative in-
memory computing applications like Decision Trees, Finite State Machines, and Hyper …

Recent self-supervised automatic speech recognition (ASR) models based on transformers
are showing best performance, but their footprint is too large to be trained on low-resource …

Most current automatic speech recognition (ASR) models use decoders that do not have
access to global contextual information at the token level. Therefore, we propose a decoder …

Self-attention (SA) is an integral part of the Transformer neural networks, originally
demonstrated its powerful ability in handling text sequences in machine translation tasks …