Automatic ear detection and feature extraction using Geometric Morphometrics and convolutional neural networks

Accurate gathering of phenotypic information is a key aspect in several subject matters, including biometrics, biomedical analysis, forensics, and many other. Automatic identification of anatomical structures of biometricinterest, such as fingerprints, iris patterns, or facial traits, are extensively used in applications like access control and anthropological research, all having in common the drawback of requiring intrusive means for acquiring the required information. In this regard, the ear structure has multiple advantages. Not only the ear's biometricmarkers can be easily captured from the distance with non intrusive methods, but also they experiment almost no changes over time, and are not influenced by facial expressions. Here we present a new method based on Geometric Morphometricsand Deep Learning for automatic ear detection and feature extraction in the form of landmarks. A convolutional neural network was trained with a set of manually landmarkedexamples. The network is able to provide morphometric landmarkson ears' images automatically, with a performance that matches human landmarking. The feasibility of using ear landmarksas feature vectors opens a novel spectrum of biometricsapplications.