Multi-modal Adversarial Training for Zero-Shot Voice Cloning

no code implementations • 28 Aug 2024 • John Janiczek, Dading Chong, Dongyang Dai, Arlo Faria, Chao Wang, Tao Wang, Yuzong Liu

The discriminator is used in a training pipeline that improves both the acoustic and prosodic features of a TTS model.

Decoder Text to Speech +1

Toward Zero Oracle Word Error Rate on the Switchboard Benchmark

no code implementations • 13 Jun 2022 • Arlo Faria, Adam Janin, Korbinian Riedhammer, Sidhi Adkoli

While commercial ASR systems are still below this threshold, a research system is shown to clearly surpass the accuracy of commercial human speech recognition.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1