Improved subword modeling for WFST-based speech recognition
Because in agglutinative languages the number of observed word forms is very high,
subword units are often utilized in speech recognition. However, the proper use of subword …
subword units are often utilized in speech recognition. However, the proper use of subword …
[PDF][PDF] Leveraging a Character, Word and Prosody Triplet for an ASR Error Robust and Agglutination Friendly Punctuation Approach.
Punctuating ASR transcript has received increasing attention recently, and well-performing
approaches were presented based on sequence-to-sequence modelling, exploiting textual …
approaches were presented based on sequence-to-sequence modelling, exploiting textual …
[PDF][PDF] User-centric evaluation of automatic punctuation in ASR closed captioning
Punctuation of ASR-produced transcripts has received increasing attention in the recent
years; RNN-based sequence modelling solutions which exploit textual and/or acoustic …
years; RNN-based sequence modelling solutions which exploit textual and/or acoustic …
Joint word-and character-level embedding CNN-RNN models for punctuation restoration
The sequence-to-sequence modelling paradigm has been successfully used in automatic
punctuation of text generated by Automatic Speech Recognizers (ASR), using bidirectional …
punctuation of text generated by Automatic Speech Recognizers (ASR), using bidirectional …
[PDF][PDF] Assessing the Semantic Space Bias Caused by ASR Error Propagation and its Effect on Spoken Document Summarization.
Ambitions in artificial intelligence involve machine understanding of human language. The
state-of-the-art approach for Spoken Language Understanding is using an Automatic …
state-of-the-art approach for Spoken Language Understanding is using an Automatic …
A prosody inspired RNN approach for punctuation of machine produced speech transcripts to improve human readability
A Moró, G Szaszák - 2017 8th IEEE International Conference …, 2017 - ieeexplore.ieee.org
Speech communication human-machine interfaces exploit automatic speech recognition to
implement speech-to-text conversion. Unfortunately, in the past, not much effort has been …
implement speech-to-text conversion. Unfortunately, in the past, not much effort has been …
[PDF][PDF] Towards abstractive summarization in Hungarian
We publish an abstractive summarizer for Hungarian, an encoder-decoder model initialized
with huBERT, and fine-tuned on the ELTE. DH corpus of former Hungarian news portals …
with huBERT, and fine-tuned on the ELTE. DH corpus of former Hungarian news portals …
[PDF][PDF] An audio-based sequential punctuation model for asr and its effect on human readability
G Szaszák - Acta Polytechnica Hungarica, 2019 - epa.niif.hu
Inserting punctuation marks into the word chain hypothesis produced by automatic speech
recognition (ASR) has long been a neglected task. In several application domains of ASR …
recognition (ASR) has long been a neglected task. In several application domains of ASR …
A bilingual comparison of maxent-and rnn-based punctuation restoration in speech transcripts
Closed captioning is a common method to improve accessibility of TV programs for people
who are hearing impaired or hard of hearing, while representing an application relevant for …
who are hearing impaired or hard of hearing, while representing an application relevant for …
Low Latency MaxEnt-and RNN-Based Word Sequence Models for Punctuation Restoration of Closed Caption Data
Abstract Automatic Speech Recognition (ASR) rarely addresses the punctuation of the
obtained transcriptions. Recently, Recurrent Neural Network (RNN) based models were …
obtained transcriptions. Recently, Recurrent Neural Network (RNN) based models were …