ETRI-Knowledge Sharing Plaform

Self-Improving Artificial Intelligence System Multi-Agent Reinforcement Learning Model-Based Dialog System Generative Models Compensation Method Deep Learning(DL) State-Of-The-Art Artificial Sample Word Embedding Models Within-Class Covariance Normalization Training Based Universal Background Model Data-Driven Approach Estimation Technique Sample Generation Shared Task Feature Dictionaries Feature Extractor Efficient Exploration Encoding Technique Image Recognition Multi-Agent System(MAS) Predator-Prey Word-Level Training Data Vector Based Conditional Mutual Information Different Levels Isolated Word Recognition Imbalance Problem Multimodal Representation N-Gram Natural Gradient Skip-Gram Fine-Tuning Channel Effect Complex Dynamics Computing Technology Embedding Method Higher Performance High Dimension Local Patches Minority Class Prediction Performance Speech Recognition System Total Variability Feature Normalization Based Prioritization Cosine Distance Channel Compensation Data Imbalance Explore Action I-Vector Learning Algorithms Large-Scale Latent Semantic Analysis(LSI) Linear Discriminant Analysis(LDA) Performance Analysis Relative Position Real-World Support Vectormachine(SVM) Spontaneous Speech Speech Assessment Starcraft Ii Automatic Speech Recognition(ASR) Best Performance Value Function Thought Model Decision Making Dictionary-Based Execution Framework Image Classification Latent Semantics Oversampling Method Proposed Model Sentence Model Sentence Similarity Active Speaker Detection Analytical Approach Context Model Clustering-Based Computer Vision(CV) Class Imbalance Data Augmentation Joint Action Imbalanced Classification Learning-Based Kullback-Leibler (Kl) Divergence Learning Curve Confidence Measure Classification Research Cognitive Computing Embedding Technique Second-Order Action Selection