Object Detection Based On Robust Similarity Metric And Deep Learning

Object detection is to locate and recognize the objects in images or videos.As object recognition accuracy can be increased by using deep learning methods,object proposal generation methods,which can achieve high performance,are becoming challenging and also has attracted much attention.Since the computational complexity of locating an object can be significantly reduced by focusing only on the generated object proposals(instead of a large number of bounding boxes with all possible positions and scales obtained by using the sliding window strategy),object proposal generation methods have been widely applied to object detection,visual tracking and other computer vision tasks.Object proposal generation methods usually have the drawbacks of low recall and high computational burden,which limit their applications.However,the huma.n visual system can quickly locate various targets,such as salient objects,aesthetic regions,general objects,biological targets,and so on.Therefore,to simulate the target localization mechanism of human visual system and construct a.fast object proposal generation method,can greatly promote the development of target detection and other computer vision tasks.In this dissertation,we make deep analysis of several representative challenges in object detection(including fast object location,low recall in video related tasks,small-sized objects,subjective objects),and we propose several methods to address these challenges.Moreover,we extend the proposed methods to several applications,such as object detection,visual tracking,saliency detection,aesthetic image cropping,face detection,and so on.The main contents and the contributions of this dissertation are as follows:(1)We propose an object discovery method via cohesion measurement.In this paper,we reinvestigate the affinity ma.trices originally used in image segmentation methods based on spectral clustering.A new affinity matrix,which is robust to color distortions,is formulated for object discovery.Moreover,a Cohesion Measurement(CM)for object regions is also derived based on the formulated affinity matrix.Based on the new Cohesion Measurement,a novel object discovery method is proposed to discover objects latent in an image by utilizing the eigenvectors of the affnity matrix.Then we apply the proposed method to both saliency detection and object proposal generation.Experimental results on several evaluation benchma.rks demonstrate that the proposed CM based method has achieved promising performa.nce for these two tasks.(2)We propose a target-specific object proposal generation method to locate objects in videos.Specifically,we propose to generate target-specific object proposals by integrating the information of two important objectness cues:colors and edges,which are complementary to each other for different challenging environments in the process of generating object proposals.As a result,the recall of the proposed TOPG method is significantly increased.Furthermore,we propose an object proposal ranking strategy to increase the rank accuracy of the generated object proposals.The proposed TOPG method has yielded significant recall gain(about 20%-60%higher)compared with several state-of-the-art object proposal methods on several challenging visual tracking datasets.Then,we apply the proposed TOPG method to the task of visual tracking and propose a TOPG-based tracker(called as TOPGT),where TOPG is used as a sample selection strategy to select a small number of high-quality target candidates from the generated object proposals.Experimental results show the superior performance of TOPGT for visual tracking compared with several other state-of-the-art visual trackers(about 3%-11%higher than the winner of the VOT2015 challenge in term of distance precision).(3)We propose a cascaded cropping regression method(CCR)to detect aesthetic regions.Image cropping,which refers to the removal of unwanted scene areas,is an important step to improve the aesthetic quality of an image.However,it is challenging to evaluate whether cropping leads to aesthetically pleasing results because the assessment is typically subjective.In this dissertation,we propose a novel cascaded cropping regression method to perform image cropping by learning knowledge from professional photographers.The proposed CCR method improves the convergence speed of the cascaded method,which directly uses random-ferns regressors.In addition,a.two-step learning strategy is proposed and used in the CCR method to address the problem of lacking labelled cropping data.Specifically,a deep convolutional neural network(CNN)classifier is first trained on large-scale visual aesthetic da.tasets.The deep CNN model is then designed to extract features from several image cropping datasets,upon which the cropping bounding boxes are predicted by the proposed CCR method.Experimental results on public image cropping datasets demonstrate that the proposed method significantly outperforms several state-of-the-art image cropping methods.(4)We propose a fast face detection method via convolutional neural network(CNN).Current face or object detection methods via CNN explicitly extract multi-scale features based on an image pyramid.However,such a strategy increases the computational burden for face detection.In this dissertation,we propose a fast face detection method based on discriminative complete features(DCFs)extracted by an elaborately designed convolutional neural network,where face detection is directly performed on the complete feature maps.DCFs have shown the ability of scale invariance,which is beneficial for face detection with high speed and promising performance.Therefore,extracting multi-scale features on an image pyramid employed in the conventional methods is not required in the proposed method,which can greatly improve its efficiency for face detection.Experimental results on several popular face detection datasets show the efficiency and the effectiveness of the proposed method for face detection.