Multi-Modal Forensic Retrieval Framework Combining Semantic Clip Embeddings with Biometric Attribute Analysis for Suspect Identification

There is an abnormal increase in the crime rate and also the number of criminals are increasing, this leads towards a great concern about the security issues. Crime preventions and criminal identification are the primary issues before the police personnel, since property and lives protection are the basic concerns of the police but to combat the crime, the availability of police personnel is limited. With the advent of security technology, cameras especially CCTV have been installed in many public and private areas to provide surveillance activities. The footage of the CCTV can be used to identify suspects on scene. This Real time criminal identification system based on face recognition works with a fully automated facial recognition system. Haar feature-based cascade classifier and OpenCV LBPH (Local Binary Pattern Histograms) Algorithms are used for Face detection and recognition. This system will be able to detect face and recognize face automatically in real time. An accurate location of the face is still a challenging task. This Framework has been widely used by researchers in order to detect the location of faces and objects in a given image. Face detection classifiers are shared by public communities, such as OpenCV. The automatically tagging feature adds a new dimension to sharing pictures among the people who are in the picture and also gives the idea to other people about who the person is in the image. In our project, we have studied and implemented a pretty simple but very effective face detection using CNN algorithm which takes human skin color into account. Our aim, which we believe we have reached, was to develop a system that can be used by police or investigation department to recognize criminal from their faces. The method of face recognition used is fast, robust, reasonably simple and accurate with a relatively simple and easy to understand algorithms and technique. Key Words: Face Detection, MTCNN, Deep Learning, Real-Time Surveillance, Criminal Identification, CNN

Facial information acquired via three-dimensional reconstruction of head computed tomography (CT) data may be considered personal information, which can be problematic for neuroimaging studies. However, no study has verified the relationship between slice thickness and face reproducibility. This study determined the relationship and match rate between image slice thickness and face detection accuracy of face-recognition software in facial reconstructed models. Head CT data of 60 cases comprising entire faces obtained under conditions of non-contrast and 1-mm slice thickness were resampled to obtain 2-10-mm slice-thickness data. Facial models, reconstructed by image thresholding, were acquired from the data. We performed face detection tests per slice thickness on the models and calculated the face detection rate. The reconstructed facial models created from 1-mm slice-thickness data and other slice thicknesses were used as training and test data, respectively. Match confidence scores were obtained via three programs, match rates were calculated per slice thickness, and generalized estimating equations were used to evaluate the match rate trend. In general, the face detection rates for the 1-10-mm slice thicknesses were 100, 100, 98.3, 98.3, 95.0, 91.7, 86.7, 78.3, 68.3, and 61.7 %, respectively. The match rates for the 2-10-mm slice thicknesses were 100, 98.3, 98.3, 95.0, 85.0, 71.7, 53.3, 28.3, and 16.7 %, respectively. The reconstructed models tended to have higher match rates as the slice thickness decreased. Thus, thin-slice head CT imaging data may increase the possibility of the information becoming personally identifiable health information.

Tracking multiple faces online in unconstrained videos is a challenging problem as faces may appear drastically different over time and identities can be inferred only based on information available from past frames. Previous tracking methods focus on face information without reference to other modality information such as a person’s overall body appearance, leading to suboptimal performance. In this paper, we propose a new online multi-face tracking method, called online multi-face tracking with multi-modality cascaded matching (OMTMCM), to improve the tracking performance by using both face and body information. The proposed OMTMCM consists of two stages, namely detection alignment and detection association. In the first stage, a detection alignment module is designed to align face detection with body detection from the same person for the subsequent detection association. In the second stage, a cascaded matching module is designed to associate face detections across frames to locate trajectory of each target face by using both face and body information. Specifically, aligned face-body detections in the current frame are matched in a cascade manner with body and face features that are selected from past frames and stored in the designed feature memory. In this way, our method can track multiple faces online with both face and body information while eliminating the possibility of face detection and body detection from the same person being separately assigned with different identities. Experimental results demonstrate our method is on par with or better than other online tracking methods for multi-face tracking.

Understanding how the human visual system recognizes objects is one of the key challenges in neuroscience. Inspired by a large body of physiological evidence, a general class of recognition models has emerged, which is based on a hierarchical organization of visual processing, with succeeding stages being sensitive to image features of increasing complexity. However, these models appear to be incompatible with some well-known psychophysical results. Prominent among these are experiments investigating recognition impairments caused by vertical inversion of images, especially those of faces. It has been reported that faces that differ 'featurally' are much easier to distinguish when inverted than those that differ 'configurally'; a finding that is difficult to reconcile with the physiological models. Here, we show that after controlling for subjects' expectations, there is no difference between 'featurally' and 'configurally' transformed faces in terms of inversion effect. This result reinforces the plausibility of simple hierarchical models of object representation and recognition in the cortex.

Real-time face detection and recognition using hybrid-information extracted from face space and facial features

The process of identifying and spotting a criminal is slow and difficult creating a criminal detection framework that could help policemen to recognize the face of a criminal or a suspect is proposed. The framework is a client-server video-based face recognition surveillance in real-time. The framework applies face detection and tracking the video footage from the camera can be used to identify suspects, criminals, runaways, missing persons, etc. This project focuses on the development of the client side of the proposed framework, face detection, and tracking using Android mobile devices. For the face detection stage, we are using Python for artificial intelligence. The face-tracking stage is based on an intelligence algorithm. The proposed face detection. Thus proposed paper explains a way to help a criminal identification system using ML and deep neural networks.

The application of face detection and recognition technology in security monitoring systems has made a huge contribution to public security. Face detection is an essential first step in many face analysis systems. In complex scenes, the accuracy of face detection would be limited because of the missing and false detection of small faces, due to image quality, face scale, light, and other factors. In this paper, a two-level face detection model called SR-YOLOv5 is proposed to address some problems of dense small faces in actual scenarios. The research first optimized the backbone and loss function of YOLOv5, which is aimed at achieving better performance in terms of mean average precision (mAP) and speed. Then, to improve face detection in blurred scenes or low-resolution situations, we integrated image superresolution technology on the detection head. In addition, some representative deep-learning algorithm based on face detection is discussed by grouping them into a few major categories, and the popular face detection benchmarks are enumerated in detail. Finally, the wider face dataset is used to train and test the SR-YOLOv5 model. Compared with multitask convolutional neural network (MTCNN), Contextual Multi-Scale Region-based CNN (CMS-RCNN), Finding Tiny Faces (HR), Single Shot Scale-invariant Face Detector (S3FD), and TinaFace algorithms, it is verified that the proposed model has higher detection precision, which is 0.7%, 0.6%, and 2.9% higher than the top one. SR-YOLOv5 can effectively use face information to accurately detect hard-to-detect face targets in complex scenes.

Most academic facilities and institutions still use the traditional methods of employee attendance recording. This is currently done manually which may be a burden on employees and takes a long time to do. Manipulation may also occur with the manual process; thus, it loses its credibility. But now, with the advent of many deep learning algorithms, we propose in this paper the use of an automated system to implement and manage the attendance recording process automatically by a face recognition technique using convolutional neural networks. Our structure is based on a modern high-precision face detection algorithm using the YOLO v4 (You Only Look Once). The Yolo v4 superior using a single GPU achieving a high speed in detecting objects comparing with other models that require to use many GPUs. Using the Darknet-53 layers for face detection we get an accuracy of 100%. The system was designed using PyQt5; an easy-use and high accuracy system which also displays attendance information in the main interface and stores it automatically in an excel file so that it can be reviewed periodically.

The complexity and applicability of facial recognition in the growing domains of security and authentication, payment, advertising and health care have accelerated the research and development of this niche in the artificial intelligence (AI) field. A facial recognition system utilizes technologies to recognize a human face through mapping of distinct facial features from raw data such as photography and video to information of known faces stored in the database. Recognizing faces in real-world images is a challenging task. Generally, a facial recognition system involves multiple processes such as facial detection, feature extraction, face matching and confidence scoring in which various techniques and algorithms are available in the form of libraries and frameworks. In this paper, an investigation is conducted using two pipelines to identify the effectiveness of two prevalent open-source libraries, OpenCV and OpenFace, on real-time face recognition system. In the OpenCV pipeline, Viola –Jones algorithm is chosen for face detection, while feature extraction is conducted using local binary pattern histogram (LBPH). Face detection is performed using dlib, while feature extraction is conducted using deep learning algorithm in the OpenFace pipeline. Performance measures of both pipelines are evaluated using Labelled Faces in the Wild (LFW) as a benchmark data set. In addition, a locally populated data set consisting of images of personnel who are given access to the office door system is used to experiment on the applicability of these pipelines on real streaming data. OpenFace achieved an AUROC of 0.98, while OpenCV scored 0.94. Both experiments inferred that facial recognition system with OpenFace pipeline outperforms OpenCV in all aspects of performance measures. Further development demonstrates the successful implementation of OpenFace pipeline to a functional prototype in replacing the manual door access control.

Face Recognition is the most popular and trending technology in the present era. It is an effective way to provide vision to a machine for better interaction with humans. The way of living will be reflected if machines can read our faces. The face recognition system will move the world in a new dimension. It will be beneficial in many ways to find the identity and security. In this paper, a face recognition system is proposed for advanced applications such as access and security, payments, criminal identifications etc. The process of identification will be based on face recognition which is further divided into three steps: detection of face, extractions of the features and classification, and real time recognition. Detection of face is recognized as the essential step of our system. It is used to extract a face in a frame, which is based on the Viola-Jones object detection algorithm that uses AdaBoost classifier with Haar and LBP features. Local Binary Patterns (LBP) is utilized to extract the unique features of the face like eyes, nose, and mouth in the feature extraction phase. The facial image is correlated with the images available in the database for the classification. The system is implemented in Python using OpenCV library. Kivy is used to create a user interface and also to build executables for different platforms.

We propose a new biologically motivated model to localize or detect faces in natural color input scene. The proposed model integrates a bottom-up saliency mechanism for extracting features from an input image and a top-down perceptual mechanism for detecting faces using the results of the bottom-up processing. For bottom-up feature extraction, we consider the roles of cells in our visual receptor for edge detection and cone opponency, and also reflect the roles of the lateral geniculate nucleus to find a symmetrical property of an interesting object such as shape and pattern. Also, independent component analysis (ICA) is used to find a filter that can generate a salient region from feature maps constructed by edge, color opponency and symmetry information, which models the role of redundancy reduction in the primary visual cortex. For the top down perceptional processing to detect faces, we partially model the role of the inferior temporal areas, which plays an important role for face recognition. Computer experimental results show that the proposed model successfully indicates faces in natural scenes.

Criminal face identification system using deep learning algorithm multi-task cascade neural network (MTCNN)

Most techniques for group emotion recognition rely on detection on faces of people and then aggregating the facial information to interpret the group emotion of a given image. However, several faces in the image may be occluded, non-frontal, or indistinguishable, i.e., too many faces in a single image (crowd). This paper focuses on such cases and investigate alternate frameworks which does not involve face detection. The developed frameworks are applied on two datasets – EmotiC and Group Affect Database 3.0 and the results are shown to be competitive with face detection (MTCNN) based approaches.

Although machine learning and artificial intelligence have been widely applied, noise and interference are still major disturbances to degrade the quality of image transmission and processing efficiency in multimedia data transmission. Based on the face detection system utilizing the machine learning algorithms and artificial intelligence, this paper examines, analyzes and compares the performances of the AdaBoost machine learning algorithm and Convolutional Neural Network (CNN)-based algorithm in processing face information, which is disturbed by channel noise and fading effect encountered in the transmission of face image. The face detection system used is based on HAAR feature extraction. The extracted HAAR features are subjected to classification training and learning of the cascade classifier. Then the face detection is performed on the picture information outside the database. Results of computer simulation show that for the image data affected by the fading and AWGN, the face recognition system still marked the positions of the eyes and mouth with high accuracy. However for fading and higher AWGN, using a machine learning algorithm with a convolutional neural network is better than the ADABOOST algorithm. It can be concluded that the machine learning algorithm can effectively reduce the adverse effect of multimedia data transmission without increasing the SNR and use of higher level of modulation scheme.

On the basis of an established individualized face 3D model, this paper implements a new facial pose detection based on face model plane. The new method, by utilizing the feature point correspondence between model plane and model image, can compute the three parameters of face pose angle and effect model display, representing a simple interactive system of virtual face pose estimation. Experiments show that proposed algorithm, simple and highly automated, can accurately and effectively detect facial pose information and display pose changes of virtual models.