Florian L. Kreyssig

2 accepted papers

2019

PyHTK: Python Library and ASR Pipelines for HTK

Chao Zhang, Florian L. Kreyssig, Qiujia Li, Philip C. Woodland

ICASSP 2019accepted

This paper describes PyHTK, which is a Python-based library and associated pipeline to facilitate the construction of large-scale complex automatic speech recognition (ASR) systems using the hidden Markov model toolkit (HTK). PyHTK can be used to generate sophisticated artificial neural network (ANN…

Cited by 0 Source ↗Scholar ↗

2018

Improved Tdnns Using Deep Kernels and Frequency Dependent Grid-RNNS

Florian L. Kreyssig, Chao Zhang, Philip C. Woodland

ICASSP 2018accepted

Time delay neural networks (TDNNs) are an effective acoustic model for large vocabulary speech recognition. The strength of the model can be attributed to its ability to effectively model long temporal contexts. However, current TDNN models are relatively shallow, which limits the modelling capabili…

Cited by 0 Source ↗Scholar ↗