ETRI-Knowledge Sharing Plaform

Automatic Translation Translation Model Artificial Intelligence System Self-Improving Barrier-Free Language Barrier Preliminary Study Recognition Model Acoustic Model Automatic Speech Recognition(ASR) Multi-Domain Contextual Information Training Data Prediction Model Speech Translation Korean Speech Neural Machine Translation Baseline Model Excellent Performance Performance Improvement Need For Conversational Data Error Reduction High-Quality Matching Pattern Model Parameter Technology Infrastructure N-Gram Text Information And Phrases Automatic Conversion Construction Procedure Code-Switching Data Sets Deep Learning(DL) European Languages Freely Available End-To-End Speech Recognition Lattice Network Lack Of Data Speech Corpus Parallel Data Parallel Corpus Time Information Over Time Arabic Speech Recognition Audio Signal Effective Communication Infrastructure Support Korean Language Model-Based Speech Recognition System Ted Talks New Method Open Technology Target Domain Translation System Source Language Clean Environment Deep Neural Network(DNN) English Language Large-Scale Parallel Spontaneous Speech Topic Model Bleu Score DNN-HMM Speech Signals Proper Names Target Language Reduction Rate F-Measure Government Open Data Knowledge-Based Lm Adaptation Speech Information Word Embedding Open-Domain Dialog Total Cost Text Corpus Correction Rate Language Model Adaptation Model Performance Text Data