Deep Learning on Binary Patterns for Face Recognition

Citations of this article
Mendeley users who have this article in their library.


In this paper an efficient and robust method for real-time face recognition is proposed. As a part of pre-processing to remove noise and unwanted features, a filter is applied to the images of standard datasets. Subsequently binary patterns are extracted from these images which are further fed into multilayer perceptron to classify images. The proposed method is tested on four benchmark datasets namely FACES94, FACES95, FACES96 and Grimace which pose challenges in illumination, pose, expression, head scale and rotation. This method delivers accuracies in the neighbourhood of 91% when tested on these datasets. Our methodology was further extended to embedded systems like Raspberry Pi 3 to give it more real-time and practical scenario. This test gave promising results and proved the model to be efficient and useful in day to day scenarios for face recognition.




Vinay, A., Gupta, A., Bharadwaj, A., Srinivasan, A., Murthy, K. N. B., & Natarajan, S. (2018). Deep Learning on Binary Patterns for Face Recognition. In Procedia Computer Science (Vol. 132, pp. 76–83). Elsevier B.V.

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free