음성감정 추출을 위한 오픈소스 기반 인공지능 설계 및 개발
Design and Development of Open-Source-Based Artificial Intelligence for Emotion Extraction from Voice
0
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
본 연구는 청각 장애인의 의사소통 개선을 목표로, 음성 데이터에서 감정을 인식하고 분류하는인공지능 모델을 개발하였다. 이를 위해 CNN-Transformer, HuBERT-Transformer, 그리고 Wav2Vec 2.0 모델을 포함하는 세 가지 주요 인공지능 모델을 활용하여, 사용자의 음성을 실시간으로 분석하고 감정을 분류한다. 음성 데이터의 특징을 효과적으로 추출하기 위해 Mel-Frequency Cepstral Coefficient(MFCC)와 같은 변환 방식을 적용, 음성의 복잡한 특성과 미묘한 감정 변화를 정확하게포착하고자 하였다. 실험 결과, HuBERT-Transformer 모델이 가장 높은 정확도를 보임으로써, 음성기반 감정 인식 분야에서의 사전 학습된 모델과 복잡한 학습 구조의 융합이 효과적임을 입증하였다. 본 연구는 음성 데이터를 통한 감정 인식 기술의 발전 가능성을 제시하며, 청각 장애인의 의사소통과 상호작용 개선에 기여할 수 있는 새로운 방안을 모색한다는 점에서 의의를 가진다.
Abstract
This study aims to improve communication for people with hearing impairments by developing artificial intelligence models that recognize and classify emotions from voice data. To achieve this, we utilized three major AI models: CNN-Transformer, HuBERT-Transformer, and Wav2Vec 2.0, to analyze users' voices in real-time and classify their emotions. To effectively extract features from voice data, we applied transformation techniques such as Mel-Frequency Cepstral Coefficient (MFCC), aiming to accurately capture the complex characteristics and subtle changes in emotions within the voice. Experimental results showed that the HuBERT-Transformer model demonstrated the highest accuracy, proving the effectiveness of combining pre-trained models and complex learning structures in the field of voice-based emotion recognition. This research presents the potential for advancements in emotion recognition technology using voice data and seeks new ways to improve communication and interaction for individuals with hearing impairments, marking its significance.
참고문헌 (19)
- S. J. Lee, "Glove-type Sign Language Translator for Communication with the Hearing-impaired Person", Journal of Rehabilitation Welfare Engineering & Assistive Technology, 17(1), pp.35~40 (2023)[DOI]
- Eunbyul Jung, "A Qualitative Study on the Communication Experience of Hearing Impairment Listeners through COVID-19 Pandemic : Using Interpretative Phenomenological Analysis", Journal of Rehabilitation Research), 27(2) (2023)[DOI]
- Hyun-sam Shin, "Deep Learning-Based Speech Emotion Recognition Technology Using Voice Feature Filters", Journal of the Korean Society of Big Data, 8(2), pp.223~231 (2023)[DOI]
- Ji-Eun Park, "Experiencing and Expression of Deaf Adolescents", KOSES, 19(3), pp.51~58 (2016)[DOI]
- Guiyoung Son, "Spontaneous Speech Emotion Recognition Based On Spectrogram With Convolutional Neural Network", The Transactions of the Korea Information Processing Society, 13(6), pp.284~290 (2024)[DOI]
- Speech recognition service, "Eunji Lee", ~, pp.ASTI MARKET INSIGHT 67 (2022)
- "Human-machine interaction (HMI), present and future"[URL]
- Jo Eun-kyoung, "A Study on Dialog Processing for Man-Machine Interaction", Hangul, 306, pp.101~130 (2014)
- Soeun Park, "Speech Emotion Recognition based on CNN using Spectrogram", Journal of Information and Control Symposium, 2018(10), pp.240~241 (2018)
- Seunghan Ha1, "Feature Extraction and Analysis of AI-Based Speech Signals for Auditory Rehabilitation", Asia-pacific Journal of Convergent Research Interchange, 9(5), pp.105~116 (2023)[DOI]
- Shikha Gupta, "Feature extraction using MFCC", Signal & Image Processing : An International Journal, 4(4), pp.101~108 (2013)
- Abdul, Zrar Kh, "Mel frequency cepstral coefficient and its applications : A review", IEEE Access, 10, pp.122136~122158 (2022)
- Ashish Vaswani, "Attention Is All You Need", Advnces in neural information processing systems (2017)
- Wei-Ning Hsu, "HuBERT : Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units", IEEE, 2021, pp.29, 3451~3460
- HeeJin Jang, "Diagnosis of Parkinson's disease based on audio voice using wav2vec", Journal of Digital Convergence, 19(12), pp.353~358 (2021)[DOI]
- Seong-Bong Yang, "Improved CNN Algorithm for Object Detection in Large Images", Journal of The Korea Society of Computer and Information, 25(1), pp.45~53 (2020)
- Ji-Seon Park, "IOptimizing CNN Structure to Improve Accuracy of Artwork Artist Classification", Journal of The Korea Society of Computer and Information, 28(9), pp.9~15 (2023)
- K. Simonyan, "Very deep convolutional networks for large-scale image recognition", Published as a conference paper at ICLR 2015 (2014)
- "DACON"[URL]