Cho, Sungjae Lee, Soo-Young 2024-01-12T04:08:33Z 2024-01-12T04:08:33Z 2022-10-04 2021 2308-457X https://pubs.kist.re.kr/handle/201004/77777 We present a methodology to train our multi-speaker emotional text-to-speech synthesizer that can express speech for 10 speakers' 7 different emotions. All silences from audio samples are removed prior to learning. This results in fast learning by our model. Curriculum learning is applied to train our model efficiently. Our model is first trained with a large single-speaker neutral dataset, and then trained with neutral speech from all speakers. Finally, our model is trained using datasets of emotional speech from all speakers. In each stage, training samples of each speaker-emotion pair have equal probability to appear in mini-batches. Through this procedure, our model can synthesize speech for all targeted speakers and emotions. Our synthesized audio sets are available on our web page. English ISCA-INT SPEECH COMMUNICATION ASSOC Multi-speaker Emotional Text-to-speech Synthesizer Conference 10.48550/arXiv.2112.03557 1 Interspeech Conference, pp.2337 - 2338 Interspeech Conference 2337 2338 FR Brno, CZECH REPUBLIC 2021-08-30 INTERSPEECH 2021 000841879502086