Journal article

Revisiting Facial Age Estimation With New Insights From Instance Space Analysis

K Smith-Miles, X Geng

IEEE Transactions on Pattern Analysis and Machine Intelligence | IEEE COMPUTER SOC | Published : 2022

Abstract

When demonstrating the effectiveness of a new algorithm, researchers are traditionally encouraged to compare their algorithm's performance against existing algorithms on well-studied benchmark test suites. In the absence of more nuanced methodologies, algorithm performance is typically summarized on average across the test suite examples. This paper highlights the potential bias of conclusions drawn by analyzing 'on average' performance, and the opportunities offered by a recent testing methodology known as instance space analysis. To illustrate, we revisit our 2007 comparative study of algorithms for facial age estimation, and rigorously stress-test to challenge the original conclusions. Th..

View full abstract

University of Melbourne Researchers

Grants

Awarded by Australian Research Council


Awarded by National Natural Science Foundation of China


Funding Acknowledgements

The authors are grateful to Dr. Mario Andres Munoz-Acosta and Dr. Neelofar (University of Melbourne) for their contributions to the development of the instance space analysis MATLAB code and the online tool MATILDA. They also thank Xi Qian (Southeast University) for providing research assistance to re-run the feature calculations and algorithms from the 2007 study. This work was funded by the Australian Research Council under Laureate Fellowship scheme (FL140100012) and the National Natural Science Foundation of China (62076063).