Papers

ExcitGlow: Improving a WaveGlow-based Neural Vocoder with Linear Prediction Analysis

International Conference
2016~2020
작성자
한혜원
작성일
2020-12-01 16:59
조회
3159
Authors : Suhyeon Oh, Hyungseob Lim, Kyungguen Byun, Min-Jae Hwang, Eunwoo Song, Hong-Goo Kang

Year : 2020

Publisher / Conference : APSIPA (*awarded Best Paper)

Research area : Speech Signal Processing, Text-to-Speech

Presentation/Publication date : 2020.12.10

Related project : Embedded Neural TTS

Presentation : Oral

In this paper we propose ExcitGlow, a vocoder that incorporates the source-filter model of voice production theory into a flow-based deep generative model. By targeting the distribution of the excitation signal instead of the speech waveform itself, we significantly reduce the size of the flow-based generative model. To further reduce the number of parameters, we apply a parameter sharing technique in which a single affine coupling layer is used for several flow layers. To avoid quality degradation, we also introduce a closed-loop training framework to optimize the flow model for both the speech and excitation signal generation processes. Specifically, we choose negative log-likelihood (NLL) loss for the excitation signal and multi-resolution spectral distance for the speech signal. As a result, we are able to reduce the model size from 87.73M to 15.60M parameters while maintaining the perceptual quality of synthesized speech.

* Awarded Best paper in APSIPA 2020

전체 370
370 International Conference Yeona Hong, Hyewon Han, Woo-jin Chung, Hong-Goo Kang "StableQuant: Layer Adaptive Post-Training Quantization for Speech Foundation Models" in IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2025
369 International Conference Sangmin Lee, Woojin Chung, Hong-Goo Kang "LAMA-UT: Language Agnostic Multilingual ASR through Orthography Unification and Language-Specific Transliteration" in Association for the Advancement of Artificial Intelligence (AAAI), 2025
368 International Journal Hyewon Han, Xiulian Peng, Doyeon Kim, Yan Lu, Hong-Goo Kang "Dual-Branch Guidance Encoder for Robust Acoustic Echo Suppression" in IEEE Transactions on Audio, Speech and Language Processing (TASLP), 2024
367 International Journal Hyungseob Lim, Jihyun Lee, Byeong Hyeon Kim, Inseon Jang, Hong-Goo Kang "Perceptual Neural Audio Coding with Modified Discrete Cosine Transform" in IEEE Journal of Special Topics in Signal Processing (JSTSP), 2025
366 International Conference Juhwan Yoon, Hyungseob Lim, Hyeonjin Cha, Hong-Goo Kang "StylebookTTS: Zero-Shot Text-to-Speech Leveraging Unsupervised Style Representation" in APSIPA ASC, 2024
365 International Conference Doyeon Kim, Yanjue Song, Nilesh Madhu, Hong-Goo Kang "Enhancing Neural Speech Embeddings for Generative Speech Models" in APSIPA ASC, 2024
364 Domestic Conference 최웅집, 김병현, 강홍구 "자기 지도 학습 특징을 활용한 음성 신호의 논 블라인드 대역폭 확장" in 대한전자공학회 2024년도 하계종합학술대회, 2024
363 Domestic Conference 홍연아, 정우진, 강홍구 "효율적인 양자화 기법을 통한 DNN 기반 화자 인식 모델 최적화" in 대한전자공학회 2024년도 하계종합학술대회, 2024
362 Domestic Conference 김병현, 강홍구, 장인선 "저지연 조건하의 심층신경망 기반 음성 압축" in 한국방송·미디어공학회 2024년 하계학술대회, 2024
361 International Conference Miseul Kim, Soo-Whan Chung, Youna Ji, Hong-Goo Kang, Min-Seok Choi "Speak in the Scene: Diffusion-based Acoustic Scene Transfer toward Immersive Speech Generation" in INTERSPEECH, 2024