A Novel Landmarked Face Database for Arab Celebrities

Document Type : Original Research Papers

Authors

1 Mathematics Dept., Faculty of Science, Benha University, Benha, Egypt

2 Computer Science Dept., Faculty of Computers and Artificial Intelligence, Benha University, Benha, Egypt

Abstract

In this paper, we present a novel database of colored and gray, plausible face images. The database contains almost 389 images of 79 Arab celebrities with automatically generated landmarks acquired from the web in wild-life. We apply face recognition (face detection) using Caffe-Model with open cv to extract faces from images then store them in 256 × 256 pixels images. The dataset is publicly available for research purposes and can be used as a training and testing material in developing various algorithms related to face detection, inpainting, frontalization, recognition, and analysis. It is challenging to inpaint face images in the wild-life, due to the large variation of appearances, such as expressions, gender, different poses, and ethnicity. This inspired us to establish a dataset that includes the faces of Arab celebrities to use it in future work.

Keywords

Files

Share

How to cite

Statistics