AutoExpress

Introduction

Automatically creates 28 different expressions from a given image using Automatic1111 Stable Diffusion WebUI API. The application uses Inpainting with After Detailer Extension to inpaint the following expressions on the face:

For the latest changes, goto: Change Log

Emotion	Emotion	Emotion	Emotion
Admiration	Amusement	Anger	Annoyance
Approval	Caring	Confusion	Curiosity
Desire	Disappointment	Disapproval	Disgust
Embarrassment	Excitement	Fear	Gratitude
Grief	Joy	Love	Nervousness
Neutral	Optimism	Pride	Realization
Relief	Remorse	Sadness	Surprise

Examples

Anime

Anime Image Emotion Grid

The above is an example anime expressions on the following image without cherry picking results.

Anime input image

Realistic

Realistic Image Emotion Grid

The above is an example realistic expressions on the following image without cherry picking results.

Realistic input image

---

Requirements

• Python3 Prefer version 10

• Automatic1111 Stable Diffusion WebUI API

  • Online Services
  • I'm currently working on integrating my CivitAI-Model-Downloader to automatically download LORAs.

• After Detailer Extension

• Additional packages for Linux

• python3-pip

• python3-venv

BUild & Run With Docker (Requires Docker)

Build

docker buildx build --tag autoexpress .

Run

docker run -dp 127.0.0.1:5000:5000 autoexpress:latest

• Clone Github Repo

git clone https://github.com/deepratnaawale/AutoExpress.git

• Goto cloned repo

cd AutoExpress

• Create and activate virtual enviorenment

python3 -m venv .venv

• WINDOWS

.venv\Scripts\activate

• LINUX

source .venv/bin/activate

• Install Requirements

pip install -r requirements.txt

• Launch App

flask --app autoexpress run

• Launch A1111 Stable Diffusion WebUI Api

You can check if the api is active by accessing the docs (if running locally) http://127.0.0.1:7860/docs#/default/

• Connect to the UI at http://127.0.0.1:5000

Usage

It's pretty Straight forward. Keep an eye on the logs in the console. Dropping an image will auto populate all attributes in the UI if the image has Stable Diffusion info text. Only supports A1111 WebUI API for now.

PS: This is a very handy tool to use with Silly Tavern to generate character expressions sprites for any bot.

Revamped UI

Future Work (Need contributers)

• Change pose along with expression keeping consistent clothes and background. (Probably via Controlnets and ADetailer)
• Better Expressions for Realistic Images.
• Support for Comfy UI, Next, etc.
• Tests to check each module.

Special Thanks

SD Parsers by d3x-at

SD API Examples

Avery Velfaire for bug fixes and prompt engineering.