Deep Learning in Speech Synthesis (Google Research Talk)
Source
Evernote/IFTTT Feedly/Deep Learning in Speech Synthesis.md
Summary
2013 년 구글 연구 발표로, 통계적 파라메트릭 음성 합성 (Statistical Parametric Speech Synthesis) 에 딥러닝을 적용한 최근 사례를 소개합니다. 기존 은닉 마르코프 모델 (HMM) 기반 접근법과 딥러닝 기반 접근법의 차이를 비교 분석합니다.
Key Points
- 통계적 파라메트릭 음성 합성 분야에 딥러닝 적용 사례 제시
- 기존 HMM 기반 방법과 딥러닝 기반 방법의 비교
- 2013 년 9 월 기준 구글 연구진 발표 자료
Related
-
Speech and Natural Language: Where Are We Now And Where Are We Headed
-
Accurate and Compact Large Vocabulary Speech Recognition on Mobile Devices
-
Speaker Adaptation of Context Dependent Deep Neural Networks
-
Language Model Verbalization for Automatic Speech Recognition
-
Target Language Adaptation of Discriminative Transfer Parsers
-
Wireless Networks Design in the Era of Deep Learning Model-Based, AI-Based, or Both
-
Coordinated Multi-Device Presentations: Ambient-Audio Identification
-
Behavioural reconfigurable and adaptive data reduction in body sensor networks
-
Improved Domain Adaptation for Statistical Machine Translation
-
Efficient Estimation of Word Representations in Vector Space
-
Scalable Decipherment for Machine Translation via Hash Sampling
-
Token and Type Constraints for Cross-Lingual Part-of-Speech Tagging
-
From mixed-mode to multiple devices. Web surveys, smartphone surveys and apps
-
Eureka: Edge-Based Discovery of Training Data for Machine Learning
-
A Semantic Matching Energy Function for Learning with Multi-relational Data
-
Continuous Birdsong Recognition Using Gaussian Mixture Modeling of Image Shape Features
-
Enlisting the Ghost: Modeling Empty Categories for Machine Translation