The Design of High-Level Features for Photo Quality Assessment

by Yan Ke, Xiaoou Tang, and Feng Jing


We propose a principled method for designing high level features for photo quality assessment. Our resulting system can classify between high quality professional photos and low quality snapshots. Instead of using the bag of low-level features approach, we first determine the perceptual factors that distinguish between professional photos and snapshots. Then, we design high level semantic features to measure the perceptual differences. We test our features on a large and diverse dataset and our system is able to achieve a classification rate of 72% on this difficult task. Since our system is able to achieve a precision of over 90% in low recall scenarios, we show excellent results in a web image search application.

A talk based on this work: Taking Great Pictures (Automatically). Guest lecture for Computational Photography. Nov 27, 2007.

Y. Ke, X. Tang, and F. Jing. In Proceedings of Computer Vision and Pattern Recognition, 2006. [PDF 800KB]

Sample Results / Supplementary Material:

Images used in ranking Ranked results (rank in filename)
Apple Apple Results
Statue of Liberty Statue of Liberty Results
Cow Cow Results
BMW BMW Results
Rose Rose Results
Violin Violin Results

Commentary on the above results.


Page maintained by Yan Ke < y k e @ c m u . e d u >

March 2006