Effects of Facial Movements to Expressive Speech Productions: A Computational Study

Conference proceedings article

ผู้เขียน/บรรณาธิการ

กลุ่มสาขาการวิจัยเชิงกลยุทธ์

ไม่พบข้อมูลที่เกี่ยวข้อง

รายละเอียดสำหรับงานพิมพ์

รายชื่อผู้แต่ง: Prom-On S., Onsri M.

ผู้เผยแพร่: Hindawi

ปีที่เผยแพร่ (ค.ศ.): 2019

หน้าแรก: 481

หน้าสุดท้าย: 484

จำนวนหน้า: 4

ISBN: 9781728101101

นอก: 0146-9428

eISSN: 1745-4557

URL: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85083257598&doi=10.1109%2fICKII46306.2019.9042701&partnerID=40&md5=8cc86c6fb1834107858c8ef5a2cc473b

ภาษา: English-Great Britain (EN-GB)

ดูบนเว็บไซต์ของสำนักพิมพ์

บทคัดย่อ

This paper presents a computational study on the relation between the movement of visual facial and acoustic features. Audio-visual corpus on expressive speech production was collected for the study. For the corpus, the relevant stimuli consist of 4 facial expression and 4 Thai sentences, in total of 16 combinations. Video and audio data were captured from 10 native Thai speakers. Each speaker pronounces the sentence with specified expression. In total, there are 160 audio-visual tracks for the analysis. Facial features were extracted and tracked by using visual markers through the pronunciation. At the same time, acoustic data, particularly the fundamental frequency (F0) was tracked and synchronized with the facial data. Computational analysis on the landmark and dynamic features of both visual and audio data were performed. This result provides the templates of the expressive facial movement together with the acoustic adjustments. ฉ 2019 IEEE.

คำสำคัญ

Audio-synchronization, Facial expression, Facial movement, Speech production