ININ Submission to Zero Cost ASR Task at MediaEval 2016
نویسندگان
چکیده
This paper details the experiments conducted to train an as good performing Vietnamese speech recognition system as possible using public domain data only, as a part of the Zero Cost task at MediEval 2016. We explored techniques related to audio preprocessing, use of speaker’s pitch information, data perturbation, for building subspace Gaussian mixture acoustic model which is known for estimating robust parameters when the amount of data is less, and also unsupervised adaptation, RNN language model based lattice rescoring and system combination using ROVER tec hnique.
منابع مشابه
Zero-Cost Speech Recognition Task at Mediaeval 2016
The main goal of the Zero-Cost Speech Recognition task is to bring researchers together on the topic of training ASR systems using only publicly available data. In particular, the task consists on the development of either an LVCSR or a subword speech recognizer on a given target language. For this year we selected Vietnamese as the target language The organizers provided participants several s...
متن کاملLACS System Analysis on Retrieval Models for the MediaEval 2014 Search and Hyperlinking Task
We describe the LACS submission to the Search sub-task of the Search and Hyperlinking Task at MediaEval 2014. Our experiments investigate how different retrieval models interact with word stemming and stopword removal. On the development data, we segment the subtitle and Automatic Speech Recognition (ASR) transcripts into fixed length time units, and examine the effect of different retrieval mo...
متن کاملDCU at MediaEval 2011: Rich Speech Retrieval
We describe our runs and results for the Rich Speech Retrieval (RSR) Task at MediaEval 2011. Our runs examine the use of alternative segmentation methods on the provided ASR transcripts to locate the beginning of the topic, assuming that this will capture or get close to the starting point of the relevant segment; combination of various types of queries and weighting of metadata to move the rel...
متن کاملUNLP at the MediaEval 2015 C@merata Task
This paper presents a description of our submission to the C@merata task in MediaEval 2015. This submission is a revision to the system submitted for the same task in MediaEval 2014 including some bug fixing. The system answers the natural language queries over the musical scores. The approach is based upon two main steps: identifying the musical entities and relations present in the query, and...
متن کاملDCU Linking Runs at MediaEval 2012 Search and Hyperlinking Task
We describe Dublin City University (DCU)’s participation in the Hyperlinking sub-task of the MediaEval 2012 Search and Hyperlinking Task. Our strategy involves combining textual metadata, automatic speech recognition (ASR) transcripts, and visual content analysis to create anchor summaries for each video segment available for linking. Two categories of fusion strategy, score-based and rank-base...
متن کامل