Emojis are an efficient way to annotate and illustrate written text in a graphic way. Some people even use them in a playful manner to substitute words. Currently, emojis are usually voiced by text-to-speech systems (TTS) by reading out a descriptive name. This is either found in data supplied by the manufacturer or taken from CLDR. These names usually differ from their official, immutable Unicode names. The idea behind this project is that the actual graphical usages of emojis better translates to non-verbal acoustic audio cues or sound clips, e.g. many face emojis relate to interjections, some of which are usable across language borders. For some emojis this seems straight-forward, for others it is more debatable.
The Clova smart speaker series by Line (at least Wave and Friends models) apparently already replace some emojis by prerecorded sound effects. http://news.livedoor.com/article/detail/15035531/ https://robotstart.info/2018/07/20/line-clova-emoji-voice.html
The first goal of this project is to create a list of corresponding sound effects for each basic emoji (i.e. no separate ones for skin tone variants, for instance, but perhaps for genders where appropriate) and the second goal is to find or create actual sound effects to be used for this.