Artist Style Classification Model Release
Person from discord launched a new AI model called Kaloscope based on arch LSNet, specifically designed to identify artists' styles.

Key Features
It can accurately recognize the styles of 31,770 artists, with an accuracy rate of 84.2%.
The model contains 100 million parameters and was trained for 80 epochs using 8 H20 GPUs, balancing performance and efficiency.The training data uses the latest dataset from Danbooru, dated October 2024.It supports real-time inference and can determine an artist's style within seconds.
Application Scenarios
Identifying the style of 2D (anime/manga) works
Automatically classifying and labeling artworksSearching and matching artists based on style
This model is open-source and free to use, ready for immediate deployment. With it, you no longer need to manually identify artists' styles—AI can help you get it done quickly.

Repo: https://github.com/spawner1145/comfyui-lsnet
To try it online: https://huggingface.co/spaces/DraconicDragon/Kaloscope-artist-style-classifier