Abstract
In this work, we present a novel visual perception-inspired local description approach as a preprocessing step for deep learning.With the ongoing growth of visual data, efficient image descriptor methods are becoming more and more important.Several local point-based description methods were defined in the past decades before the highly accurate and popular deeplearning methods such as convolutional neural networks (CNNs) emerged. The method presented in this work combines anovel local description approach inspired by the Gestalt laws with deep learning, and thereby, it benefits from both worlds.To test our method, we conducted several experiments on different datasets of various forensic application domains, e.g.,makeup-robust face recognition. Our results show that the proposed approach is robust against overfitting and only littleimage information is necessary to classify the image content with high accuracy. Furthermore, we compared our experimentalresults to state-of-the-art description methods and found that our method is highly competitive. For example it outperformsa conventional CNN in terms of accuracy in the domain of makeup-robust face recognition.
Reference
Hörhan, M., & Eidenberger, H. (2021). Gestalt descriptions for deep image understanding. Pattern Analysis and Applications, 24(1), 89–107. https://doi.org/10.1007/s10044-020-00904-6