Implementation of Face Detection System Based on PNX1500 Platform

HU Dao-yuan,ZHANG Li

2009-01-01

Abstract:A face detection system based on Adaboost algorithm is implemented on the PNX1500 platform,and the system can detect the fronted face correctly and stably.With optimization tool,the detection speed can reach up 5 frames per second with CIF(352×288) videos.Furthermore,the system can meet the requirement of detection and tracking application.

What problem does this paper attempt to address?

An Intelligent Surveillance System Base on Accelerated Human and Face Detection

song xiaofeng,xie di,chen deren,zhang weize,tong ruofeng

DOI: https://doi.org/10.1007/978-3-642-03718-4_84

2012-01-01

Abstract:In this paper, we propose a system to acquire high resolution face image of a human in surveillance scene. Firstly, the system detects a full or partially occluded human body in the scene and zooms in to it automatically. And then a face detection and zooming process is running. Finally, a high resolution face image is captured and saved. The system can run real-time as the algorithms are accelerated by multi-core CPU and CPU. Experiments show that the system can capture most high resolution face image of human in the scene and can be implemented in the real surveillance systems. Our system provides an alternative method with better result to acquire high resolution face image comparing to face super-resolution methods.
Real-time Detection and Tracking Algorithm Based on the Color and Character of Human Face

XIANG Gui-shan,WANG Xuan-yin,LIANG Dong-tai

DOI: https://doi.org/10.3969/j.issn.1003-501x.2007.04.010

2007-01-01

Abstract:For fast detecting and tracking human face in real-time video, a method based on the color and character of face was proposed. Moving human-body objects were extracted from video according to temporal differencing and shape information firstly, and then the skin area was segmented based on its Gaussian model established due to the clustering character of skin color in YCbCr color space. After noise was suppressed, the non-face skin area was eliminated according to the geometric characters of face and the color characters of organs in face, and then the face was located in image. According to the face location, the head and shoulder part were centered in video by driving camera using control strategy, so movement tracking was realized. In order to enhance the adaptability for light changes, proper updating strategy for the Gaussion model parameters was proposed. Experiments show that the proposed method performs well in face detection and tracking in both simple background and complex background.
A Novel Multi-User Face Detection under Infrared Illumination by Real Adaboost

Yan Chao,Wang Yuanqing

DOI: https://doi.org/10.1109/cise.2009.5366152

2009-01-01

Abstract:Location and tracking the human faces is one of the critical technologies in free stereoscopic display system. But because of illumination variation or facial expression, it is difficult to detect human faces accurately and fast. In this paper, an infrared face detection based on real Adaboost algorithm and Cascade structure is implemented. With active infrared illumination, the problem caused by variation of illumination is almost solved, which makes the detection system more robust. Meanwhile, the combination of real Adaboost and Cascade structure pays more attention to the human faces which is more difficult to identify, making the detection quicker a lot. In the detection of video sequence, all human faces can be detected, and misdetection rarely appears. The average processing time on a windows XP, PIV 2.4 GHz PC is about 40 ms for a 640*480-pixel image. So the improved detection is real-time. In addition, experiment proves that the improved detection behaves better when there is variation of facial expression or a little degree leaning of human face. Obviously, the improved detection is more efficient and could be used widely.
ASFD: Automatic and Scalable Face Detector

Jian Li,Bin Zhang,Yabiao Wang,Ying Tai,Zhenyu Zhang,Chengjie Wang,Jilin Li,Xiaoming Huang,Yili Xia

DOI: https://doi.org/10.1145/3474085.3475372

2022-01-01

Abstract:Along with current multi-scale based detectors, Feature Aggregation and Enhancement (FAE) modules have shown superior performance gains for cutting-edge object detection. However, these hand-crafted FAE modules show inconsistent improvements on face detection, which is mainly due to the significant distribution difference between its training and applying corpus, i.e. COCO vs. WIDER Face. To tackle this problem, we essentially analyse the effect of data distribution, and consequently propose to search an effective FAE architecture, termed AutoFAE by a differentiable architecture search, which outperforms all existing FAE modules in face detection with a considerable margin. Upon the found AutoFAE and existing backbones, a supernet is further built and trained, which automatically obtains a family of detectors under the different complexity constraints. Extensive experiments conducted on popular benchmarks, i.e. WIDER Face and FDDB, demonstrate the state-of-the-art performance-efficiency trade-off for the proposed automatic and scalable face detector (ASFD) family. In particular, our strong ASFD-D6 outperforms the best competitor with AP 96.7/96.2/92.1 on WIDER Face test, and the lightweight ASFD-D0 costs about 3.1 ms, i.e. more than 320 FPS, on the V100 GPU with VGA-resolution images.
Design and implementation of a surveillance camera system with face recognition functionality

Hu, R.,Jing Zhang,Huimin Yu,Yunhai Liu

DOI: https://doi.org/10.1109/EDSSC.2014.7061234

2014-01-01

Abstract:In this paper, we present a surveillance camera system with face recognition functionality. The face detection algorithm is based on AdaBoost and Haar-like features, as proposed in [1]. The recognition process is an extension of the Local Binary Pattern (LBP) algorithm which we have proposed in [2]. The system utilizes a digital signal processor (DSP) to implement the detection and recognition algorithms. An ARM processor is also used to provide network layer interfaces to connect the system with computers and other digital devices. Thus, we have built a low-cost, time-efficient and accurate face recognition system for surveillance cameras.
Real-time Hardware Face Detection Based on Adaboost Algorithm

TANG Qi,SU Guang-da

DOI: https://doi.org/10.3969/j.issn.1000-3428.2008.07.088

2008-01-01

Abstract:This paper presents a hardware implementation of face detection system that can detect any number of faces with any size at any position in a 256×256 image, at a speed of 35 fps when the FPGA works at 70 MHz. The detection system can achieve a detection rate of 85% with a false alarm rate of 1.5×10-6. To achieve high speed face detection, three contributions are made in the hardware architecture: real-time hardware implementation of the calculations of integral image and integral squared image, deep pipeline architecture of weak classifier implementation, and parallel multi-memory organization.
Face Detection Method for Real-Time Anti-Interference

ZHOU Jin,WANG Yuan-qing,FAN Ke-feng

DOI: https://doi.org/10.3969/j.issn.1000-7024.2013.04.050

2013-01-01

Abstract:A robust face detection method for real-time video stream is proposed.In the front stage of the system,pyramid structure,integral image and skin-color coverage rate are combined to obtain candidate areas;Cascade classifiers can detect faces effectively,and the way to select the smoothing factor in real AdaBoost algorithm is improved to accelerate convergence and avoid overfitting,meanwhile reasonable sample distribution and weights of weak classifiers insure low false alarm rate and robustness,and parallel computing based on GPU can accelerate training remarkably;In the rear stage,Kalman filter is applied for prediction and denoising,and the combination of tracking and detection speeds up processing.The experimental results show that this method can detect faces fast and accurately,and can effectively overcome various interferences such as environmental factors,poses,expressions and occlusions.
An AdaBoost-Based Face Detection System Using Parallel Configurable Architecture with Optimized Computation

Shouyi Yin,Peng Ouyang,Xu Dai,Leibo Liu,Shaojun Wei

DOI: https://doi.org/10.1109/jsyst.2015.2418680

IF: 4.802

2017-01-01

IEEE Systems Journal

Abstract:With the development of image sensor technology, the AdaBoost-based face detections are widely used in many monitoring sensor networks and mobile-camera-based applications. Fast face detection with high detection accuracy and low power consumption in such kinds of applications is very important. Since the AdaBoost-based face detection exhibits characteristics of data computation in dual direction and data diversity, we propose an AdaBoost-based face detection system using parallel configurable architecture with optimized computation. The architecture consists of parallel configurable arrays and two-level shared memory systems. It achieves dual-direction-based integral image computation that improves parallel processing efficiency and enables the subwindow adaptive cascade classification for data diversity, which further improves the detection efficiency in diverse face detection. Compared with the state-of-the-art works, this work achieves maximal performance of 30 frames/s at 1080p detection speed and extreme low power consumption.
P-FAD: Real-time Face Detection Scheme on Embedded Smart Cameras

Qiang Wang,Jing Wu,Chengnian Long,Bo Li

DOI: https://doi.org/10.1109/jetcas.2013.2256823

IF: 5.877

2013-01-01

IEEE Journal on Emerging and Selected Topics in Circuits and Systems

Abstract:Face detection is known to be a resource-hungry process in computer vision. Most of existing detection algorithms, such as Viola-Jones detector, are too time-consuming to be implemented on resource-limited embedded smart cameras. Observing that the computation increases proportionally to its pixel manipulation, and inspired from the cascade structure which imposes more complex processing on the promising regions, hierarchical approach may be a feasible solution. The challenge issue in the design is how to construct a multi-layer architecture, in which the complex processing can be split from the pixel manipulation and guarantee detection accuracy simultaneously. To remedy this problem, we propose a novel Pyramid-like FAce Detection (P-FAD) scheme that results in a significant reduction of computation during face detection process. P-FAD has five-layer architecture, in which the operating units decrease dramatically from top to down while imposing complex computations in the last layer. We have implemented our scheme both on a notebook and our embedded smart camera platform. Our experimental results demonstrate that the P-FAD schemes resource-aware properties while still hold the acceptable detection accuracy.
Face Detection Method Research and Implementation Based on Adaboost

Dai Chen,Ji-lin Wang,Yong Zhou

DOI: https://doi.org/10.1109/iptc.2010.125

2010-01-01

Abstract:This paper describes and deduces the theory of Haar-like features, Integral image and AdaBoost algorithm, which were proposed by Paul Viola, and then researches its improvement. We combine Microsoft Visual C++6.0 with OpenCV Function library to develop the software, and achieve the function of real-time face detection. According to experimental results, we can conclude that the improved algorithm can detect the real-time face better, having higher success rate and lower false alarm rate. Meanwhile, with analyzing the experimental process and the results, this paper proposes using Skin-Color model to integrate the AdaBoost algorithm, and then using subsidiary discriminant function to improve the detection process.
Fast AdaBoost-Based Face Detection System on a Dynamically Coarse Grain Reconfigurable Architecture.

Jian Xiao,Jinguo Zhang,Min Zhu,Jun Yang,Longxing Shi

DOI: https://doi.org/10.1587/transinf.e95.d.392

2012-01-01

IEICE Transactions on Information and Systems

Abstract:An AdaBoost-based face detection system is proposed, on a Coarse Grain Reconfigurable Architecture (CGRA) named "REMUS-II". Our work is quite distinguished from previous ones in three aspects. First, a new hardware-software partition method is proposed and the whole face detection system is divided into several parallel tasks implemented on two Reconfigurable Processing Units (RPU) and one micro Processors Unit (mu PU) according to their relationships. These tasks communicate with each other by a mailbox mechanism. Second, a strong classifier is treated as a smallest phase of the detection system, and every phase needs to be executed by these tasks in order. A phase of Haar classifier is dynamically mapped onto a Reconfigurable Cell Array (RCA) only when needed, and it's quite different from traditional Field Programmable Gate Array (FPGA) methods in which all the classifiers are fabricated statically. Third, optimized data and configuration word pre-fetch mechanisms are employed to improve the whole system performance. Implementation results show that our approach under 200 MHz clock rate can process up-to 17 frames per second on VGA size images, and the detection rate is over 95%. Our system consumes 194 mW, and the die size of fabricated chip is 23 mm(2) using TSMC 65 nm standard cell based technology. To the best of our knowledge, this work is the first implementation of the cascade Haar classifier algorithm on a dynamically CGRA platform presented in the literature.
Infrared Face Detection Based on Real Adaboost Algorithm and Cascade Structure

严超,王元庆

DOI: https://doi.org/10.3969/j.issn.1001-5078.2009.11.027

2009-01-01

Abstract:Location and tracking the human faces is one of the critical technologies in free stereoscopic display system.But because of illumination variation and some other reasons,it is difficult to detect human faces accurately and fast.In this paper,an infrared face detection based on real Adaboost algorithm and Cascade structure is implemented.With active infrared illumination and separating of visible light,the problem caused by variation of illumination is almost solved.Meanwhile,the combination of real Adaboost and Cascade structure pays more attention to the human faces which is more difficult to identify,making the detection more robust and quicker a lot.In the detection of video sequence,all human faces can be detected,and misdetection rarely appears.The average processing time on a windows XP,PIV 2.4 GHz PC is about 40 ms for a 640×480-pixel image.So the improved detection is real-time.In addition,experiment proves that the improved detection is robust when there is variation of facial expression or a little degree leaning of human face.
Implementation of Face Detecting Algorithm of AdaBoost Based on DSP

金友芝,苏光大,魏良

2008-01-01

Abstract:To solve the technical problem of real-time face detection,DSP is used to implement the algorithm of face detection.The AdaBoost algorithm of human face detection is optimized.A novel feature definition is proposed and an improved cascade-structured classifier which is called fuzzy cascade classifier is presented.The DSP chip of TI Company and its periphery circuit are introduced.The whole procedure of the detecting system is described.Finally,the DSP program is optimized using the CCS.The detecting speed reaches 26 films per second with 256× 256 pixels image input.The error rate is only 0.9%.The real-time face detection is implemented.
Adaboost based Real-time Pedestrian Detection System

Zhu Yiqiang,Zhang Hongcai,Cheng Yongmei,Yang Tao,Zhao Chunhui

DOI: https://doi.org/10.3321/j.issn:1671-4598.2006.11.010

2006-01-01

Abstract:To solve the problem of pedestrian detection in far-fie ld intelligent video surveillance system, this paper provides a detection system b ased on object's appearance characteristics. Four types of rectangle filters whi ch efficiently describe the local characteristics of people's appearance are giv en and some modified mothed to adaboost algorithm are introduced, which decrease the computation time and optimize the system configuration. The result of exper iments show the system has a high detection rate and realize real-time processin g. The common noise disturbing which exist in the surveillance scenes such as sh adows, raining, leaves blowing in the wind have little effect on the detection r esults, the system runs the 320 240 pixel video image sequence at above 15 frame s/second on a 3.0 GHz P4 processor.
Real-time Face Detection by Adaboost Based on Infrared Illumination

钟翔,王元庆

DOI: https://doi.org/10.3969/j.issn.1000-3428.2009.08.075

2009-01-01

Abstract:This paper proposes a face locating method based on active infrared illumination and Adaboost, which greatly improves performance in robustness and speed. In this system, an active infrared illumination structure is used to get eye candidates, and then face candidates are extracted, followed by a cascaded real Adaboost algorithm to detect all faces. Experimental result shows that it takes no more than 6 ms to locate all faces in one frame for a video stream with a resolution of 640×480.
A Face Detection Based on Frontal-Inverse Adaboost Algorithm

YANG Cheng-ping,WANG Yuan-qing

2008-01-01

Abstract:Accurately locating and tracking the position of the user's face is important for autostereoscopic display. In this paper, a cascade face detector learnt by using real Adaboost algorithm is presented, and for the purpose of reducing the false alarm, a novel procedure called Frontal-Inverse Adaboost which detects the sub-window twice( one is frontal and another is inverse) is proposed. A set of experiments are presented, the results indicate that the precision improving to normal Adaboost algorithm is satisfying. It can be seen that the proposed method is very efficient and has lots of significant values to autostereoscopic display.
An Algorithm of Fast Face Detection in Video Based on AdaBoost

Yafeng DENG,Guangda SU,Bo FU

DOI: https://doi.org/10.3969/j.issn.1000-3428.2006.11.081

2006-01-01

Abstract:A fast face detection algorithm in video integrated with tracking is designed.It uses a cascade-structured face detector trained with AdaBoost to detect face.Integrated with a linear face tracker,searching region and rescaling levels of the detector are reduced,and thus detection speed is accelerated.It also designs a classifier training method integrated with AdaBoost,bootstrapping and weight-trimming,which can be easily used in object detection.
Implementation of human eye detection system Based on DM642

ZHENG Wei,WANG Yuan-qing

DOI: https://doi.org/10.3969/j.issn.1004-373X.2012.04.034

2012-01-01

Abstract:A design of real-time eye detection system based on TI DM642 DSP is proposed.The detecting algorithm combines eye′s bright pupil effect under active illumination."Face-Eye" two level structure algorithm with Adaboost and SVM is adopted to search and locate user′s eyes.Meanwhile,the algorithm is optimized according to DSP′s characters.The experiment results indicate that this DSP eye detection system can reach the requirement of real-time processing for 640*480 pixels continuous video,and the accuracy and stability are satisfactory.
Efficient improvement for adaboost based face detection system

PuFeng Wu,Hongzhi Liu,Xixin Cao,Jing Liu,Zhonghai wu

2009-01-01

Abstract:The training of the adaboost algorithm for face detection is time costly; it often needs days or weeks in the previous system. In this paper, we describe efficient optimization techniques and implement skills to reduce the training time. First we use some preprocessing technique to reduce the candidate features size to ten percent of the original, and then we use some implement skills to further reduce the training time. Besides these, we use double thresholds to describe each feature, which can improve the efficient of each feature, and reduce the required feature number for the final strong classifier. The experiment result show that the training of our system is hundred time faster than previous systems.
A Multi-View Face Detection Based on Real Adaboost Algorithm

Wu Bo,Huang Chang,Ai Haizhou,Lao Shihong

DOI: https://doi.org/10.1360/crad20050924

2005-01-01

Abstract:In this paper, a multi-view face detection method based on real Adaboost algorithm is presented. Human faces are divided into several viewpoint categories according to their poses in 3D, and for each of these categories we design a form of weak classifiers in look-up-table (LUT) type using Haar-like features that have confidences in real values as their outputs, and correspondingly construct its space of weak classifiers from which the cascade face detector is learnt by using real Adaboost algorithm. For speed up, multi-resolution searching and pose prediction strategies are introduced. For frontal face detection the experiments on CMU+MIT frontal face test set result a 94.5% correct rate with 54 false alarms; for multi-view face detection the experiments on CMU profile face test set result a 89.8% correct with 221 false alarms. The average processing time on a PIV 2.4GHz PC is about 80 ms for a 320×240-pixel image. It can be seen that the proposed method is very efficient and has significant value in application.

Implementation of Face Detection System Based on PNX1500 Platform

An Intelligent Surveillance System Base on Accelerated Human and Face Detection

Real-time Detection and Tracking Algorithm Based on the Color and Character of Human Face

A Novel Multi-User Face Detection under Infrared Illumination by Real Adaboost

ASFD: Automatic and Scalable Face Detector

Design and implementation of a surveillance camera system with face recognition functionality

Real-time Hardware Face Detection Based on Adaboost Algorithm

Face Detection Method for Real-Time Anti-Interference

An AdaBoost-Based Face Detection System Using Parallel Configurable Architecture with Optimized Computation

P-FAD: Real-time Face Detection Scheme on Embedded Smart Cameras

Face Detection Method Research and Implementation Based on Adaboost

Fast AdaBoost-Based Face Detection System on a Dynamically Coarse Grain Reconfigurable Architecture.

Infrared Face Detection Based on Real Adaboost Algorithm and Cascade Structure

Implementation of Face Detecting Algorithm of AdaBoost Based on DSP

Adaboost based Real-time Pedestrian Detection System

Real-time Face Detection by Adaboost Based on Infrared Illumination

A Face Detection Based on Frontal-Inverse Adaboost Algorithm

An Algorithm of Fast Face Detection in Video Based on AdaBoost

Implementation of human eye detection system Based on DM642

Efficient improvement for adaboost based face detection system

A Multi-View Face Detection Based on Real Adaboost Algorithm