Steam コミュニティ :: 動画 :: Introducing Tio - The VR AI Bot [VRChat AI Bot

ストアページ

VRChat

すべてスレッドスクリーンショット作品ブロードキャスト動画ニュースガイドレビュー

VRChat > 動画 > Cneq のムービー

Introducing Tio - The VR AI Bot [VRChat AI Bot - Tio]

VRChat

すべてのムービーを表示

アワード

タイトル

説明

Hey guys.

First off I'd like to give thanks to Skuld for demonstrating that VR AI bots are possible. If I didn't happen to see Joken in the great pug all the time I most likely wouldn't of created my own.

Here is a showcase of my first project, a VR AI Bot called Tio.

I started working on this bot last month and after getting everything working and also going through the process of creating an avatar for it It's in a state that I'm happy with.

There are tons of improvements I want to add to it such as integration with the VRChat API and I also want to eventually replace IBM Watson TTS [temporarily] to test and showcase what the future may look like when bots have near perfect human speech [which is possible with 11labs].

While I was working on the bot I've also been training a lot of voices with 11labs and have that process down to a T.

I will most likely upload a showcase of the voices I have cloned and also a tutorial on the process to clone a voice [since most people seem to be gatekeeping how to do it].

Besides that around 5 days ago I started working with stable diffusion and I'm very impressed with how far AI art has come and I will also upload some content about that which will include showcases of my art and also some tutorials/recommendations for people wanting to get into stable diffusion.

Long-term I have plans on integrating all of this AI content into a VR world that will both showcase what is possible with AI and also teach in an interactive way [world is already created I'm just too lazy to start working in unity again lol].

Now, despite all these plans my university classes start on march 2nd and I'm taking some pretty hard units so I will most likely be busy and not able to work on much but rest assured it it will come eventually lol.

Thanks guys.

PS-

For video length purposes the footage has been trimmed slightly, when interacting with Tio you can expect to see a further 5-10 seconds between what you ask Tio and when Tio replies.

This is largely due to IBM Watson TTS being extremely slow and when it comes to recognizing speech with Whisper AI and GPT-3 generating a response it is extremely fast [I can see this in the command line] yet the TTS takes more than double that amount of time to be generated.

セーブキャンセル

243	ユニーク訪問者数
1	現在のお気に入り数

2 件のコメント

< >

Your_White_Knight 2024年10月16日 23時42分

Interesting.... and a bit scary, heh

Lucas Gintoki 2023年2月24日 0時21分

damn I really made friends with you by getting trolled by your bot :()I

< >