A System for Simultaneous Translation of Lectures and Speeches

@inproceedings{Fgen2009ASF,
  title={A System for Simultaneous Translation of Lectures and Speeches},
  author={Christian F{\"u}gen},
  year={2009}
}
This thesis realizes the first existing automatic system for simultaneous speech-to-speech translation. The focus of this system is the automatic translation of (technical oriented) lectures and speeches from English to Spanish, but the different aspects described in this thesis will also be helpful for developing simultaneous translation systems for other domains or languages. 
Towards the evaluation of simultaneous speech translation from a communicative perspective
TLDR
A comparison of machine speech-to-speech and speech- to-text translation systems with respect to human performances in similar communicative tasks and how the performance of such systems is perceived by final users is presented. Expand
A real-world system for simultaneous translation of German lectures
TLDR
A real-time automatic speech translation system for university lectures that can interpret several lectures in parallel that is now being installed in several lecture halls at KIT and is able to provide the translation to the students in several parallel sessions. Expand
Consecutive Decoding for Speech-to-text Translation
TLDR
This work proposes COnSecutive Transcription and Translation (COSTT), an integral framework for speech-to-text translation that outperforms the previous state-of-the-art methods. Expand
A Corpus of Spontaneous Speech in Lectures: The KIT Lecture Corpus for Spoken Language Processing and Translation
TLDR
This paper describes how the disfluencies in the data are annotated and provides detailed statistics on the size of the corpus and the speakers and the results of the translation of a source text without different sorts of dis fluencies or no disfluency at all. Expand
The KIT Lecture Corpus for Speech Translation
TLDR
The corpus of German lectures at KIT should be suited for training a spoken language translation system the traditional way, but should also enable the translation system to automatically and autonomously adapt itself to the varying topics and speakers of lectures. Expand
A hybrid phonotactic language identification system with an SVM back-end for simultaneous lecture translation
TLDR
A hybrid of both approaches that clearly outperforms the PPR system while not adding any additional computing time is built, which is the choice for the use in the lecture translation system due to its faster run-time and good performance. Expand
Speech Recognition Machine Translation Speech Synthesis Acoustic Model Language ModelDictionary Language Model Translation Model Voice
In this work, we investigate methods to automatically adapt our simultaneous lecture translation systems to the diverse topics that occur in educational lectures. Utilizing materials that areExpand
SDST: Successive Decoding for Speech-to-text Translation
TLDR
SDST is proposed, an integral framework with Successive Decoding for end-to-end speech- to-text translation and improves the previous state-of-the-art methods by big margins. Expand
Low Latency ASR for Simultaneous Speech Translation
TLDR
In order to minimize the latency, a combination of run-on decoding with a technique for identifying stable partial hypotheses when stream decoding and a protocol for dynamic output update that allows to revise the most recent parts of the transcription is combined. Expand
Unsupervised Vocabulary Selection for Domain-Independent Simultaneous Lecture Translation
TLDR
A novel approach based on feature-based ranking and evaluate the effectiveness of 21 different features and their combinations for this task significantly improved vocabulary coverage, reducing out-of-vocabulary rate and reducing the latency of the end-to-end system. Expand
...
1
2
3
...

References

SHOWING 1-10 OF 165 REFERENCES
Open Domain Speech Translation: From Seminars and Speeches to Lectures
TLDR
The paper presents the speech translation performance of these systems on lectures and gives an overview of the final real-time lecture translation system. Expand
Open Domain Speech Recognition & Translation:Lectures and Speeches
  • C. Fügen, M. Kolss, +4 authors A. Waibel
  • Computer Science
  • 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings
  • 2006
TLDR
Starting from a translation system for European parliamentary plenary sessions and a lecture speech recognition system it is shown how both components perform in unison on speech translation of lectures. Expand
The ISL Baseline Lecture Transcription System for the TED Corpus
This paper describes the Interactive Systems Laboratories’ automatic lecture transcription system for the Translanguage English Database (TED) corpus, which provides text-hypothesis for theExpand
Cross domain automatic transcription on the TC-STAR EPPS corpus
TLDR
The ongoing development of the British English European Parliament Plenary Session corpus will be part of the speech-to-speech translation evaluation infrastructure of the European TC-STAR project and first recognition results on the English speech recordings are presented. Expand
Speech Translation Enhanced ASR for European Parliament Speeches - On the Influence of ASR Performance on Speech Translation
TLDR
By porting an STE-ASR framework to the task at hand, it is shown how the word errors for transcribing English and Spanish speeches can be lowered by 3.0% and 4.8% relative, respectively. Expand
Input Segmentation of Spontaneous Speech in JANUS: A Speech-to-speech Translation System
TLDR
This paper describes how multi-level segmentation of single utterance turns improves translation quality and facilitates accurate translation in the Janus speech-to-speech translation system. Expand
The influence of utterance chunking on machine translation performance
TLDR
This paper addresses the question how utterance chunking influences machine translation performance in an empirical study and compares different relatively simple chunking/ segmentation strategies on speech recognition hypotheses as well as on reference transcripts. Expand
The use of prosody in a combined system for punctuation generation and speech recognition
TLDR
A combined system for punctuation generation and speech recognition that incorporates prosodic information with acoustic and language model information is discussed, which can improve the Fmeasure of punctuation recognition by 19% relative. Expand
The ISL TC-STAR Spring 2006 ASR Evaluation Systems
The project Technology and Corpora for Speech to Speech Translation (TC-STAR) aims at making a break-through in speech-to-speech translation research, significantly reducing the gap between theExpand
Language model adaptation using WFST-based speaking-style translation
  • Takaaki Hori, D. Willett, Y. Minami
  • Computer Science
  • 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03).
  • 2003
TLDR
A weighted finite-state transducer (WFST) that translates sentence styles from in-domain to out-of-domain and enables to integrate language models of different styles of speaking or dialects and even of different vocabularies. Expand
...
1
2
3
4
5
...