Hardware Accelerators for Real-Time Face Recognition: A Survey

Abstract

Real-time face recognition has been of great interest in the last decade due to its wide and variant critical applications which include biometrics, security in public places, and identification in login systems. This has encouraged researchers to design fast and accurate embedded and portable systems that are capable to detect and recognize a large number of faces at almost video frame rate. Due to the increasing volume of reference faces, traditional general purpose computing engines such as the ones based on Intel's Pentium processors have shown not to be adequate and various dedicated hardware accelerators based on either Graphical Processing Unit (GPU), Field Programmable Gate Arrays (FPGA), Application Specific Integrated Circuit (ASIC), or even multi-core Central Processing Units (CPU) have emerged. Earlier published review papers on face detection/recognition have discussed face detection and face recognition algorithms enhancement that improve the detection accuracy. Nevertheless, none of them has reviewed the hardware accelerators used for this application. Accordingly, this paper aims to provide a comprehensive review of the most recent face recognition algorithms and associated embedded hardware systems targeting real-time performance. A detailed comparison between neural network and non-neural network-based algorithms in terms of accuracy and processing time is provided. Discussions on their suitability to be implemented into parallel hardware architectures such as Single Instruction Multiple Thread (SIMT) or Single Instruction Multiple Data (SIMD) is also discussed.

Modern Trends in Improving the Technical Characteristics of Devices and Systems for Digital Image Processing

“…Work [4] is devoted to hardware accelerators for real-time face recognition. A lot of algorithms and their implementations based on CPU, GPU, and FPGA are analyzed to solve problems from the area under consideration.…”

“…Many have one or more desktop or laptop computers and other digital devices with even more tools for creating and distributing visual content. In addition to everyday life, digital imaging and image processing devices have been introduced into various industries, medical diagnostics, satellite systems, are actively used by law enforcement agencies, and so on [3,4]. The characteristics of digital images are constantly improving including spatial resolution and color depth.…”

A Survey of Trusted Computing Solutions Using FPGAs

“…FPGA is a programmable semiconductor device with many logic gates, input/output blocks, and programmable routing that connects these logic gates, lookup tables (LUTs), and block-on-chip memory (BRAM). The connection between these blocks is called configurable logic blocks (CLB), or floating-point digital signal processing (DSP) [16]. Therefore, it can be programmed to process different data types according to its configuration.…”

“…Nowadays, FPGAs are considered a powerful device to deploy different applications using a hardware architecture described as intellectual property (IP), which means processing, memory, control, and communication that can connect multiple IPs to increase scalability. Due to FPGA' functionalities, mainly in processing large workloads efficiently, they are found in applications in the following areas: 5G open Radio Access Network (RAN) [20], prototype design [21], machine learning [22], deep learning [23], reinforcement learning [24], real-time face recognition [16], unmanned aerial vehicle control [25], multi-signal processing [26] and related to this review, trusted execution environments [27].…”

AutoFace: How to Obtain Mobile Neural Network-Based Facial Feature Extractor in Less Than 10 Minutes?

“…Various mobile and edge devices often have significantly different processing capabilities [12], [13], making it challenging to create a single facial descriptor that can effectively perform real-time face analysis with high accuracy across all devices [14]. Thus, existing frameworks have a set of predefined descriptors (small, medium, large, etc.)…”