ETRI-Knowledge Sharing Plaform

Acoustic Model Self-Improving Korean Speech Deep Neural Network(DNN) Robust Speech Recognition Artificial Intelligence System Signal Noise Ratio(SNR) Speech Database Automatic Speech Recognition(ASR) Data Augmentation Speech Recognition System State-Of-The-Art Speech Features Transformer-Based Bidirectional Lstm Assessment Method Augmentation Techniques Audio Data Conversational Speech Recognition Car Navigation System Context-Sensitive Barrier-Free Truncated Singular Value Decomposition Viterbi Decoding Endpoint Detection Algorithm Embedded Devices Human-Robot Interface Low-Resource Nonlinear Manifold Objective Function Performance Improvement Parameter Reduction Response Latency Automatic Translation Decision Fusion User-Perceived Latency High Performance Localization Method Model Parameter Multiple Speakers Noisy Environments Server-Client Connectionist Temporal Classification(CTC) Cross-Modality Blind Source Separation(BSS) Deep Learning(DL) Voice Activity Detector End-To-End Speech Recognition Field Environment Linear Regression Limited Resources Long Sequence Post-Processing Method Reduction Technique Real-World Applications Real-Time Operation Room Acoustic Signal-To-Noise Augmentation Method Computing Technology Factorization Approach Fine-Tuning Non-Stationary Noise Robot System Separation Device Channel Selection Beam Search Frame Skipping Localization Algorithm Matrix Factorization Manifold Structure Network-Based Overlapped Speech Detection Phone Recognition Proficiency Assessment Art Performance Error Rate Far-Field Low-Rank Matrix Low Latency Online Speech Recognition Recognition Model Sound Source Localization Acoustic Features Current State Digital Signal Processing Bidirectional Long Short-Term Memory Based Approach Target Matrix Euclidean Norm System Accuracy Audio-Visual Audio And Video Computational Complexity Computation Method Cognitive Computing Language Barrier Manifold Regularized Non-Native Speakers Noise Separation Speech Feature Extraction