A data-driven method for input feature selection within neural prosody generation

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

Volumn 1, Issue , 2002, Pages

  Erdem, Caglayan ^a   Zimmermann, Hanz Georg ^a  

^a DRESDEN UNIVERSITY OF TECHNOLOGY   (Germany)

Author keywords

[No Author keywords available]

Indexed keywords

ACOUSTIC PROSODY GENERATION MODELING; NEURAL NETWORK ARCHITECTURE;

LEARNING SYSTEMS; LINGUISTICS; MATHEMATICAL MODELS; MULTILAYER NEURAL NETWORKS; SPEECH ANALYSIS;

SPEECH SYNTHESIS;

EID: 0036299196     PISSN: 15206149     EISSN: None     Source Type: Conference Proceeding    
DOI: None     Document Type: Conference Paper

References (10)

1. F0 generation with a database of natural F0 patterns and with a neural network
2. Optimisation of a neural network for pitch contour generation
3. Natural f0-contours with a new neural-network-hybrid-approach
4. Early stopping - But when?
5. Prosody generation with a neural network: Weighing the importance of input parameters
6. Data-driven importance analysis of linguistic and phonetic information
7. A multi-lingual system for the determination of phonetic word stress using soft feature selection by neural networks
8. F how to train neural networks
9. Parametric description of F0-contours in a prosodic database