subpic / ava-mlsp Public

Notifications You must be signed in to change notification settings
Fork 16
Star 77

MLSP feature learning for the AVA aesthetics database

77 stars 16 forks Branches Tags Activity

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 10 Commits
metadata		metadata
models		models
.gitattributes		.gitattributes
LICENSE		LICENSE
README.md		README.md
extract_mlsp.ipynb		extract_mlsp.ipynb
predict_mlsp_wide.ipynb		predict_mlsp_wide.ipynb
train_mlsp_narrow.ipynb		train_mlsp_narrow.ipynb
train_mlsp_narrow_aug.ipynb		train_mlsp_narrow_aug.ipynb
train_mlsp_wide.ipynb		train_mlsp_wide.ipynb

Repository files navigation

ava-mlsp

This is part of the code for the paper "Effective Aesthetics Prediction with Multi-level Spatially Pooled Features"

Multi-level Spatially-Pooled (MLSP) features are used to train predictors on the AVA aesthetics database.

The code shows how to train models based on both narrow and wide MLSP features. Several models trained on AVA are included.

About

MLSP feature learning for the AVA aesthetics database

Report repository

Releases

No releases published

Packages

No packages published

Languages

Jupyter Notebook 100.0%