STT

STT: Fast, Lean, and Ubiquitous

Covers how our STT can transform your applications by enabling client-side, low-latency, and privacy-preserving speech recognition capabilities.

Read
TTS

Double Decoder Consistency

Describes Double Decoder Consistency (DDC) a novel technique to fight against TTS attention based alignment problems.

Read
STT

Streaming RNNs in TensorFlow

Describes how the STT engine’s architecture was changed to allow for streaming, achieving real-time transcription performance.

Read
TTS

Better Tacotron Alignment

Introduces two methods that improve attention alignment in Tacotron models.

Read
STT

A Journey to <10% Word Error Rate

Announces the initial release of our open source speech recognition model so that anyone can develop compelling speech experiences.

Read
TTS

Gradual Training

Introduces a gradual training scheme to ease the training difficulty in Tacotron models.

Read