ETRI-Knowledge Sharing Plaform

Self-Improving Artificial Intelligence System Reduction Rate Automatic Speech Recognition(ASR) Machine Translation(MT) Matrix Decomposition Speaker Adaptation Speech Corpus Training Data End-To-End Speech Recognition Model-Based Open Data Conventional Methods Clean Environment Deep Neural Network(DNN) Large-Scale New Method Transformation Method Target Domain Singular Value Decomposition(SVD) Spontaneous Speech Affine Transformation Large Vocabulary Continuous Speech Recognition Computationally Expensive Recognition Model Lm Adaptation Correction Rate Open-Domain Dialog Recognition Performance Language Model Adaptation Deep Learning(DL) Code-Switching Adaptive Parameter Error Reduction Performance Levels Linear Transformation Online Services