R
Ramya Rasipuram
Researcher at Idiap Research Institute
Publications - 32
Citations - 451
Ramya Rasipuram is an academic researcher from Idiap Research Institute. The author has contributed to research in topics: Hidden Markov model & Lexicon. The author has an hindex of 13, co-authored 31 publications receiving 386 citations. Previous affiliations of Ramya Rasipuram include École Polytechnique Fédérale de Lausanne & Apple Inc..
Papers
More filters
Proceedings ArticleDOI
Siri On-Device Deep Learning-Guided Unit Selection Text-to-Speech System.
Tim Capes,Paul Coles,Alistair Conkie,Ladan Golipour,Abie Hadjitarkhani,Qiong Hu,Nancy Huddleston,Melvyn J. Hunt,Jiangchuan Li,Matthias Neeracher,Kishore Prahallad,Tuomo Raitio,Ramya Rasipuram,Greg Townsend,Becci Williamson,David A. Winarsky,Zhizheng Wu,Hepeng Zhang +17 more
TL;DR: Apple’s hybrid unit selection speech synthesis system, which provides the voices for Siri with the requirement of naturalness, personality and expressivity, is described and various techniques that enable on-device capability such as preselection optimization, caching for low latency, and unit pruning for low footprint are described.
Proceedings ArticleDOI
Controllable Neural Text-to-Speech Synthesis Using Intuitive Prosodic Features.
TL;DR: The authors trained a sequence-to-sequence neural network conditioned on acoustic speech features to learn a latent prosody space with intuitive and meaningful dimensions, and showed that a model conditioned on sentence-wise pitch, pitch range, phone duration, energy, and spectral tilt can effectively control each prosodic dimension and generate a wide variety of speaking styles.
Proceedings Article
Grapheme-based Automatic Speech Recognition using KL-HMM
TL;DR: This work presents a novel grapheme-based ASR system that jointly models phoneme and graphe me information using Kullback-Leibler divergence-based HMM system (KL-HMM).
Proceedings ArticleDOI
Fast and flexible Kullback-Leibler divergence based acoustic modeling for non-native speech recognition
TL;DR: This paper investigates an approach that addresses the challenge of acoustic variability present in multi-accented non-native speech with limited amount of training data by using Kullback-Leibler divergence based hidden Markov models (KL-HMM).
Journal ArticleDOI
Articulatory feature based continuous speech recognition using probabilistic lexical modeling
TL;DR: Analysis of the probabilistic relationship captured by the parameters has shown that the approach is capable of adapting the knowledge-based phoneme-to-AF representations using speech data; and allows different AFs to evolve asynchronously.