Search Results - "Klejch, Ondrej"

Adaptation Algorithms for Neural Network-Based Speech Recognition: An Overview by Bell, Peter, Fainberg, Joachim, Klejch, Ondrej, Li, Jinyu, Renals, Steve, Swietojanski, Pawel

“…We present a structured overview of adaptation algorithms for neural network-based speech recognition, considering both hybrid hidden Markov model / neural…”
Get full text

Sequence-to-sequence models for punctuated transcription combining lexical and acoustic features by Klejch, Ondrej, Bell, Peter, Renals, Steve

“…In this paper we present an extension of our previously described neural machine translation based system for punctuated transcription. This extension allows…”
Get full text

Punctuated transcription of multi-genre broadcasts using acoustic and lexical approaches by Klejch, Ondrej, Bell, Peter, Renals, Steve

“…In this paper we investigate the punctuated transcription of multi-genre broadcast media. We examine four systems, three of which are based on lexical…”
Get full text

Ava Active Speaker: An Audio-Visual Dataset for Active Speaker Detection by Roth, Joseph, Chaudhuri, Sourish, Klejch, Ondrej, Marvin, Radhika, Gallagher, Andrew, Kaver, Liat, Ramaswamy, Sharadh, Stopczynski, Arkadiusz, Schmid, Cordelia, Xi, Zhonghua, Pantofaru, Caroline

“…Active speaker detection is an important component in video analysis algorithms for applications such as speaker diarization, video re-targeting for meetings,…”
Get full text

Towards Zero-Shot Code-Switched Speech Recognition by Yan, Brian, Wiesner, Matthew, Klejch, Ondrej, Jyothi, Preethi, Watanabe, Shinji

“…In this work, we seek to build effective code-switched (CS) automatic speech recognition systems (ASR) under the zero-shot set-ting where no transcribed CS…”
Get full text

The Edinburgh International Accents of English Corpus: Towards the Democratization of English ASR by Sanabria, Ramon, Bogoychev, Nikolay, Markl, Nina, Carmantini, Andrea, Klejch, Ondrej, Bell, Peter

“…English is the most widely spoken language in the world, used daily by millions of people as a first or second language in many different contexts. As a…”
Get full text

Efficient Intelligibility Evaluation Using Keyword Spotting: A Study on Audio-Visual Speech Enhancement by Valentini-Botinhao, Cassia, Aldana Blanco, Andrea Lorena, Klejch, Ondrej, Bell, Peter

“…We propose a new method for human speech intelligibility evaluation based on keyword spotting. In this method, participants play a stimulus and select the word…”
Get full text

Speech Collage: Code-Switched Audio Generation by Collaging Monolingual Corpora by Hussein, Amir, Zeinali, Dorsa, Klejch, Ondrej, Wiesner, Matthew, Yan, Brian, Chowdhury, Shammur, Ali, Ahmed, Watanabe, Shinji, Khudanpur, Sanjeev

“…Designing effective automatic speech recognition (ASR) systems for Code-Switching (CS) often depends on the availability of the transcribed CS resources. To…”
Get full text

Beyond Oversmoothing: Evaluating DDPM and MSE for Scalable Speech Synthesis in ASR by Minixhofer, Christoph, Klejch, Ondrej, Bell, Peter

“…Synthetically generated speech has rapidly approached human levels of naturalness. However, the paradox remains that ASR systems, when trained on TTS output…”
Get full text

Exploring Dominant Paths in CTC-Like ASR Models: Unraveling the Effectiveness of Viterbi Decoding by Zhao, Zeyu, Bell, Peter, Klejch, Ondrej

“…Connectionist Temporal Classification (CTC) has emerged as a fundamental technique in Automatic Speech Recognition (ASR), renowned for its ability to…”
Get full text

Speaker Adaptive Training Using Model Agnostic Meta-Learning by Klejch, Ondrej, Fainberg, Joachim, Bell, Peter, Renals, Steve

“…Speaker adaptive training (SAT) of neural network acoustic models learns models in a way that makes them more suitable for adaptation to test conditions…”
Get full text

Deciphering Speech: a Zero-Resource Approach to Cross-Lingual Transfer in ASR by Klejch, Ondrej, Wallington, Electra, Bell, Peter

“…We present a method for cross-lingual training an ASR system using absolutely no transcribed training data from the target language, and with no phonetic…”
Get full text

Acoustic Word Embeddings for Untranscribed Target Languages with Continued Pretraining and Learned Pooling by Sanabria, Ramon, Klejch, Ondrej, Tang, Hao, Goldwater, Sharon

“…Acoustic word embeddings are typically created by training a pooling function using pairs of word-like units. For unsupervised systems, these are mined using…”
Get full text

ASR and Emotional Speech: A Word-Level Investigation of the Mutual Impact of Speech and Emotion Recognition by Li, Yuanchao, Zhao, Zeyu, Klejch, Ondrej, Bell, Peter, Lai, Catherine

“…In Speech Emotion Recognition (SER), textual data is often used alongside audio signals to address their inherent variability. However, the reliance on human…”
Get full text

Towards Zero-Shot Code-Switched Speech Recognition by Yan, Brian, Wiesner, Matthew, Klejch, Ondrej, Jyothi, Preethi, Watanabe, Shinji

“…In this work, we seek to build effective code-switched (CS) automatic speech recognition systems (ASR) under the zero-shot setting where no transcribed CS…”
Get full text

Acoustic Model Adaptation from Raw Waveforms with Sincnet by Fainberg, Joachim, Klejch, Ondrej, Loweimi, Erfan, Bell, Peter, Renals, Steve

“…Raw waveform acoustic modelling has recently gained interest due to neural networks' ability to learn feature extraction, and the potential for finding better…”
Get full text

AVSE Challenge: Audio-Visual Speech Enhancement Challenge by Blanco, Andrea Lorena Aldana, Valentini-Botinhao, Cassia, Klejch, Ondrej, Gogate, Mandar, Dashtipour, Kia, Hussain, Amir, Bell, Peter

“…Audio-visual speech enhancement is the task of improving the quality of a speech signal when video of the speaker is available. It opens-up the opportunity of…”
Get full text

The Edinburgh International Accents of English Corpus: Towards the Democratization of English ASR by Sanabria, Ramon, Bogoychev, Nikolay, Markl, Nina, Carmantini, Andrea, Klejch, Ondrej, Bell, Peter

“…English is the most widely spoken language in the world, used daily by millions of people as a first or second language in many different contexts. As a…”
Get full text

Hierarchical recurrent neural network for story segmentation using fusion of lexical and acoustic features by Tsunoo, Emiru, Klejch, Ondrej, Bell, Peter, Renals, Steve

“…A broadcast news stream consists of a number of stories and it is an important task to find the boundaries of stories automatically in news analysis. We…”
Get full text

Lattice-Based Unsupervised Test-Time Adaptation of Neural Network Acoustic Models by Klejch, Ondrej, Fainberg, Joachim, Bell, Peter, Renals, Steve

“…Acoustic model adaptation to unseen test recordings aims to reduce the mismatch between training and testing conditions. Most adaptation schemes for neural…”
Get full text