Large Scale Image Aesthetics Evaluation
-
Updated
Jan 28, 2022 - Jupyter Notebook
Large Scale Image Aesthetics Evaluation
Official Code for Assessing UHD Image Quality from Aesthetics, Distortions, and Saliency
1. Perceptron: The very basic entity in Machine Learning. It's training and weights update in code. 2. Image Aesthetic Assessment: Determining the aesthetic content of an image. The network defined use Spatial Pyramid Pooling. 3. Image Classification: Alexnet architecture in Keras for image classification. Find more here
Self-hosted, ready-to-use API server for OCR, background removal, and additional features.
Local, private AI photo culling for travel photographers — score, group faces, search, and polish thousands of shots offline. No cloud, no account.
Web annotation platform for collecting personalized image aesthetic ratings — 142 raters, ~98k ratings across art, fashion, and scenery. Produced the source data for the XPASS-Vis dataset.
ClipCrop: Conditioned Cropping Driven by Vision-Language Model Conference
Per-attribute diagnosis of how photographic aesthetics transfers (and fails) to art — linear probes on frozen DINOv2/CLIP/SigLIP2.
To associate your repository with the image-aesthetics topic, visit your repo's landing page and select "manage topics."