Attention model for articulatory features detection
This work addresses speech-related tasks such as pronunciation training and TTS, but it appears incremental as it builds on existing LAS architecture with a new decoding technique.
The paper tackled articulatory features detection by applying the Listen, Attend and Spell architecture to phone recognition on small datasets like TIMIT and introduced a novel decoding technique for end-to-end training of articulatory detectors. They achieved results in joint phone recognition and articulatory features detection using multitask learning, though no concrete numbers are provided.
Articulatory distinctive features, as well as phonetic transcription, play important role in speech-related tasks: computer-assisted pronunciation training, text-to-speech conversion (TTS), studying speech production mechanisms, speech recognition for low-resourced languages. End-to-end approaches to speech-related tasks got a lot of traction in recent years. We apply Listen, Attend and Spell~(LAS)~\cite{Chan-LAS2016} architecture to phones recognition on a small small training set, like TIMIT~\cite{TIMIT-1992}. Also, we introduce a novel decoding technique that allows to train manners and places of articulation detectors end-to-end using attention models. We also explore joint phones recognition and articulatory features detection in multitask learning setting.