The deep convolutional neural networks for face recognition, from DeepFace to the recent FaceNet, demand a sufficiently large volume of filters for feature extraction, in addition to being deep. The shallow filter-bank approaches, e.g., principal component analysis network (PCANet), binarized statistical image features (BSIF), and other analogous variants, endure the filter scarcity problem that not all PCA and ICA filters available are discriminative to abstract noise-free features. This paper extends our previous work on multi-fold filter convolution (â.,3-FFC), where the pre-learned PCA and ICA filter sets are exponentially diversified by â.,3 folds to instantiate PCA, ICA, and PCA-ICA offspring. The experimental results unveil that the 2-FFC operation solves the filter scarcity state. The 2-FFC descriptors are also evidenced to be superior to that of PCANet, BSIF, and other face descriptors, in terms of rank-1 identification rate (%).
|Title of host publication||Second International Workshop on Pattern Recognition|
|Editors||Guojian Chen, Xudong Jiang, Masayuki Arai|
|Publication status||Published - 2017|
|Event||2nd International Workshop on Pattern Recognition, IWPR 2017 - Singapore, Singapore|
Duration: 2017 May 1 → 2017 May 3
|Name||Proceedings of SPIE - The International Society for Optical Engineering|
|Other||2nd International Workshop on Pattern Recognition, IWPR 2017|
|Period||17/5/1 → 17/5/3|
Bibliographical noteFunding Information:
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (NO. 2016R1A2B4011656).
© COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.
All Science Journal Classification (ASJC) codes
- Electronic, Optical and Magnetic Materials
- Condensed Matter Physics
- Computer Science Applications
- Applied Mathematics
- Electrical and Electronic Engineering