[4Yin-B] ポスター7

[4Yin-B-KA03] 研究会優秀賞：「アフレコ時の声優の音響特性と表情変化 —プロとノービスの比較調査—」言語・音声理解と対話処理研究会（SIG-SLUD）

〇武久 浩柊¹、藤野 千広¹、関根 和生¹ (1. 早稲田大学)

Voice acting is a representative profession that supports Japan’s anime industry. Previous research on professional voice actors has identified distinctive acoustic patterns through speech analysis. At the same time, it is well established that acoustic properties of speech are closely linked to changes in facial expressions and body movements. However, within existing voice-actor research, few studies have examined speech acoustics and facial expressions simultaneously under controlled recording conditions. To address this limitation, the present study conducted a laboratory experiment in which recording conditions were standardized across participants. We compared professional voice actors with acting-inexperienced individuals (novices) to characterize (1) acoustic features of voice-acted speech and (2) concurrent facial expression changes during dubbing performance. The results showed that professional voice actors produced a larger pitch range than novices, as evidenced by a greater difference between maximum and minimum fundamental frequency. In addition, voice actors spoke with higher vocal intensity (louder voice) than novices. These findings indicate that professional voice actors exhibit distinctive acoustic characteristics relative to novices. Regarding facial expressions, professional voice actors displayed emotion-consistent facial movements during dubbing, suggesting that they modulate facial expressions in accordance with the intended affect while performing. Overall, by jointly assessing acoustic measures and facial expression dynamics under controlled conditions, this study provides empirical evidence for multimodal features that differentiate professional voice acting from novice performance.