Onoma-to-wave: Environmental Sound Synthesis from Onomatopoeic Words

In this paper, we propose a new framework for environmental sound synthesis using onomatopoeic words and event labels. The conventional method of synthesis, in which only labels are used, cannot finely control the time-frequency structural features synthesized sounds, such as duration, timbre, pitch. There various ways to express other than labels, use words. An word, is character sequence phonetically imitating sound, has been shown be effective describing phonetic feature sounds. We believe that will enable us fine thus from on basis sequence-to-sequence framework. To convert also structure frequency property Our subjective experiments show proposed achieves same level quality WaveNet. Moreover, our methods better terms expressiveness sounds