Browsing by Author "Karasolak, Mustafa"

Now showing 1 - 2 of 2

Face Photograph Recognition via Generation from Sketches using Convolutional Neural Networks
(2019) Karasolak, Mustafa; Choupani, Roya
Face photo-sketch matching is an important problem for law enforcement agencies in terms of identifying suspects. In this study, a new sketch-photo generation and recognition technique is proposed by using residual convolutional neural network architecture. The suggested RCNN architecture consists of 6 convolutions, 6 ReLU, 4 poolings, 2 deconvolution layers. The proposed architecture is trained with face photos and sketches. Sketches are supplied as an input to the RCNN architecture and, generated face photos are obtained as the output. Then, the generated face photos are compared with the photos of the people in the database. Structural Similarity Index (SSIM) is used to measure the pairwise similarity and the photo with the highest index score is matched. CUHK Face Sketch Database containing 188 images is tested. In the experiments, 148, 20, and 20 images are used for training, validation, and testing, respectively. Data augmentation applied to 148 training images produced 444 images. Experimental results show that the success of the training curve is 90.55% and the validation success is 91.1%. True face recognition success from generated face images with SSIM is 93.89% for CUHK Face Sketch database (CUFS) and 84.55% AR database.
Matching composite drawings and mugshot photographs to determine the identity of the person
(2019) Karasolak, Mustafa
In this thesis, a new photo-sketch generation and recognition technique is proposed using residual convolutional neural network architecture. For this, the proposed architecture is trained with face photos and sketches. Sketches are applied to the proposed Region-based Convolutional Neural Networks (RCNN) architecture and, face photos are obtained at network output. Then, the obtained face photographs are compared with the images in the database. It is associated with the highest similarity photograph. Structural Similarity Index (SSIM) is used to measure similarity. It is very useful for law enforcement for image processing applications. 188 images are used for training and testing. Of these, 148 are used for training. 20 are used for validation and 20 are used for testing. Data augmentation is applied to 148 images used for training. As a result of the data augmentation process, 444 face images are obtained and used for network training. As a result of network training, the success of the training curve is 90.55% and the validation success is 91.1%. True face recognition success from generated face images with SSIM is 93.89% for CUHK database and 84.55% AR database.