Fast and Accurate 3D Face Recognition

Abstract

In this paper we present a new robust approach for 3D face registration to an intrinsic coordinate system of the face. The intrinsic coordinate system is defined by the vertical symmetry plane through the nose, the tip of the nose and the slope of the bridge of the nose. In addition, we propose a 3D face classifier based on the fusion of many dependent region classifiers for overlapping face regions. The region classifiers use PCA-LDA for feature extraction and the likelihood ratio as a matching score. Fusion is realised using straightforward majority voting for the identification scenario. For verification, a voting approach is used as well and the decision is defined by comparing the number of votes to a threshold. Using the proposed registration method combined with a classifier consisting of 60 fused region classifiers we obtain a 99.0% identification rate on the all vs first identification test of the FRGC v2 data. A verification rate of 94.6% at FAR = 0.1% was obtained for the all vs all verification test on the FRGC v2 data using fusion of 120 region classifiers. The first is the highest reported performance and the second is in the top-5 of best performing systems on these tests. In addition, our approach is much faster than other methods, taking only 2.5 seconds per image for registration and less than 0.1 ms per comparison. Because we apply feature extraction using PCA and LDA, the resulting template size is also very small: 6 kB for 60 region classifiers.

Face 3D biometrics goes mobile: Searching for applications of portable depth sensor in face recognition

“…Another way of analyzing the spatial face surface consists of finding a set of characteristics points. In recent articles, there are also mixed algorithms which combine well-known 2D face recognition algorithms with the processing of 3D features [4].…”

“…The faces are first cropped using a sphere of arbitrarily defined radius, like in [4] and [7]. It was decided to use radius 100mm.…”

Comparing strategies for 3D face recognition from a 3D sensor

“…The pixelwise comparison of two 2.5D images does not require any explicit indexing [21] and thus the matching is much faster than a matching between two 3D point clouds. These 2.5D images registered with the multiple references are sufficient to establish identity, without further feature extractions as shown in [14] [16]. We further divide the 2.5D facial images into different facial areas (the nose, eyes region, cheeks 2 We use 4 canonical faces in all experiments.…”

Twins 3D face recognition challenge

“…Algorithm 1 is a variation of McKeon. The major differ ence is the preprocessing step where the face is first roughly aligned using the symmetry plane estimation method de scribed in Spreeuwers [13], and the image is then aligned to a reference face using ICP.…”