KeyFrame extraction based on face quality measurement and convolutional neural network for efficient face recognition in videos

Improved likelihood ratios for face recognition in surveillance video by multimodal feature pairing

“…To avoid processing a whole video, keyframe extraction methods for face recognition in videos have been developed. Abed et al [ 5 ] propose a method based on face quality and deep learning. The first step is face detection using the MTCNN detector, which detects five landmarks (the eyes, the two corners of the mouth, and the nose) and then limits face boundaries to a bounding box and from there provides a confidence score.…”

“…The issue with some existing methods of face image quality computation is their dependence on subjective or indirect measures of quality, which may not necessarily align with the needs of face recognition systems. In contrast, these newer methods, as exemplified by the works of Abed et al [ 5 ] and Bahroun et al [ 7 ], provide a more direct measure of face image quality, which is closely tied to the performance of the face recognition model itself.…”

“…With the proliferation of surveillance cameras and the capture of images of criminal events, the comparison of faces has become a key tool for gathering intelligence, guiding investigations, and providing evidence in court [ 1 , 2 ]. While deep-learning based FR methods have demonstrated strong recognition performance for still images [ 3 ], such as those in the Labeled Faces in the Wild (LFW) dataset [ 4 ], video-based FR has not been as widely developed by the research community [ 5 ]. Video FR, however, offers additional information, such as temporal details and multiple views on the same person, which can be used in conjunction with frame based face recognition techniques to quickly identify subjects of interest in CCTV footage [ 6 ].…”

Feature Extraction from Several Angular Faces Using a Deep Learning Based Fusion Technique for Face Recognition

“…This article employs the correlation coefficient similarity criterion to compare the estimated feature vectors with the feature vectors of the images in the dataset. The correlation coefficient is calculated using the following formula: 𝐧 𝐦 (9) where X is the estimated feature vector and Y is the feature vector of the image in the dataset. Also, X ̅ is the mean of X, and Y ̅ is the mean of Y.…”

“…In other words, as the angle of the face to the camera increases, the accuracy of face recognition methods *Corresponding Author Email: h.hassanpour@shahroodut.ac.ir (H. Hassanpour) decreases. Recently, feature fusion technique [9,10] has improved the performance of facial recognition systems to some extent. In a facial recognition system, the fusion of information can be done at the decision level or at the feature level [11].…”