UMD Theses and Dissertations
Permanent URI for this collectionhttp://hdl.handle.net/1903/3
New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a given thesis/dissertation in DRUM.
More information is available at Theses and Dissertations at University of Maryland Libraries.
Browse
4 results
Search Results
Item Efficient Detection of Objects and Faces with Deep Learning(2020) Najibi, Mahyar; Davis, Larry S.; Computer Science; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Object detection is a fundamental problem in computer vision and is an essential building block for many applications such as autonomous driving, visual search, and object tracking. Given its large-scale and real-time applications, scalable training and fast inference are critical. Deep neural networks, although powerful in visual recognition, can be computationally expensive. Besides, they introduce shortcomings such as lack of scale-invariance and inaccurate predictions in crowded scenes that can affect detection. This dissertation studies the intrinsic problems which emerge when deep convolutional neural networks are used for object and face detection. We introduce methods to overcome these issues which are not only accurate but also efficient. First, we focus on the problem of lack of scale-invariance. Performing inference on a multi-scale image pyramid, although effective, increases computation noticeably. Moreover, multi-scale inference really blooms when the model is also trained using expensive multi-scale approaches. As a result, we start by introducing an efficient multi-scale training algorithm called "SNIPER" (Scale Normalization for Image Pyramids with Efficient Re-sampling). Based on the ground-truth annotations, SNIPER sparsely samples high-resolution image regions wherever needed. In contrast to training, at inference, there is no ground-truth information to guide region sampling. Thus, we propose "AutoFocus". AutoFocus predicts regions to be zoomed-in from low resolutions at inference time, making it possible to skip a large portion of the input pyramid. While being as efficient as single-scale detectors, these methods boost performance noticeably. Second, we study the problem of efficient face detection. Compared to generic objects, faces are rigid and crowded scenes containing hundreds of faces with extreme scales are more common. In this dissertation, we present "SSH" (Single Stage Headless Face Detector). A method that unlike two-stage localization/classification detectors, performs both tasks in a single stage, efficiently models scale variation by design, and removes most of the parameters from its underlying network, but still achieves state-of-the-art results on challenging benchmarks. Furthermore, for the two-stage detection paradigm, we introduce "FA-RPN" (Floating Anchor Region Proposal Network). FA-RPN takes the spatial structure of faces into account and allows modification of the prediction density during inference to efficiently deal with crowded scenes. Finally, we turn our attention to the first step in two-stage localization/classification detectors. While neural networks were deployed for classification, localization was previously solved using classic algorithms which became the bottleneck. To remedy, we propose "G-CNN" which models localization as a search in the space of all possible bounding boxes and deploys the same neural network used for classification. Furthermore, for tasks such as saliency detection, where the number of predictions is typically small, we develop an alternative approach that runs at speeds close to 120 frames/second.Item Towards a Fast and Accurate Face Recognition System from Deep Representations(2019) Ranjan, Rajeev; Chellappa, Rama; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The key components of a machine perception algorithm are feature extraction followed by classification or regression. The features representing the input data should have the following desirable properties: 1) they should contain the discriminative information required for accurate classification, 2) they should be robust and adaptive to several variations in the input data due to illumination, translation/rotation, resolution, and input noise, 3) they should lie on a simple manifold for easy classification or regression. Over the years, researchers have come up with various hand crafted techniques to extract meaningful features. However, these features do not perform well for data collected in unconstrained settings due to large variations in appearance and other nuisance factors. Recent developments in deep convolutional neural networks (DCNNs) have shown impressive performance improvements in various machine perception tasks such as object detection and recognition. DCNNs are highly non-linear regressors because of the presence of hierarchical convolutional layers with non-linear activation. Unlike the hand crafted features, DCNNs learn the feature extraction and feature classification/regression modules from the data itself in an end-to-end fashion. This enables the DCNNs to be robust to variations present in the data and at the same time improve their discriminative ability. Ever-increasing computation power and availability of large datasets have led to significant performance gains from DCNNs. However, these developments in deep learning are not directly applicable to the face analysis tasks due to large variations in illumination, resolution, viewpoint, and attributes of faces acquired in unconstrained settings. In this dissertation, we address this issue by developing efficient DCNN architectures and loss functions for multiple face analysis tasks such as face detection, pose estimation, landmarks localization, and face recognition from unconstrained images and videos. In the first part of this dissertation, we present two face detection algorithms based on deep pyramidal features. The first face detector, called DP2MFD, utilizes the concepts of deformable parts model (DPM) in the context of deep learning. It is able to detect faces of various sizes and poses in unconstrained conditions. It reduces the gap in training and testing of DPM on deep features by adding a normalization layer to the DCNN. The second face detector, called Deep Pyramid Single Shot Face Detector (DPSSD), is fast and capable of detecting faces with large scale variations (especially tiny faces). It makes use of the inbuilt pyramidal hierarchy present in a DCNN, instead of creating an image pyramid. Extensive experiments on publicly available unconstrained face detection datasets show that both these face detectors are able to capture the meaningful structure of faces and perform significantly better than many traditional face detection algorithms. In the second part of this dissertation, we present two algorithms for simultaneous face detection, landmarks localization, pose estimation and gender recognition using DCNNs. The first method called, HyperFace, fuses the intermediate layers of a DCNN using a separate CNN followed by a multi-task learning algorithm that operates on the fused features. The second approach extends HyperFace to incorporate additional tasks of face verification, age estimation, and smile detection, in All-In-One Face. HyperFace and All-In-One Face exploit the synergy among the tasks which improves individual performances. In the third part of this dissertation, we focus on improving the task of face verification by designing a novel loss function that maximizes the inter-class distance and minimizes the intraclass distance in the feature space. We propose a new loss function, called Crystal Loss, that adds an L2-constraint to the feature descriptors which restricts them to lie on a hypersphere of a fixed radius. This module can be easily implemented using existing deep learning frameworks. We show that integrating this simple step in the training pipeline significantly boosts the performance of face verification. We additionally describe a deep learning pipeline for unconstrained face identification and verification which achieves state-of-the-art performance on several benchmark datasets. We provide the design details of the various modules involved in automatic face recognition: face detection, landmark localization and alignment, and face identification/verification. We present experimental results for end-to-end face verification and identification on IARPA Janus Benchmarks A, B and C (IJB-A, IJB-B, IJB-C), and the Janus Challenge Set 5 (CS5). Though DCNNs have surpassed human-level performance on tasks such as object classification and face verification, they can easily be fooled by adversarial attacks. These attacks add a small perturbation to the input image that causes the network to mis-classify the sample. In the final part of this dissertation, we focus on safeguarding the DCNNs and neutralizing adversarial attacks by compact feature learning. In particular, we show that learning features in a closed and bounded space improves the robustness of the network. We explore the effect of Crystal Loss, that enforces compactness in the learned features, thus resulting in enhanced robustness to adversarial perturbations. Additionally, we propose compact convolution, a novel method of convolution that when incorporated in conventional CNNs improves their robustness. Compact convolution ensures feature compactness at every layer such that they are bounded and close to each other. Extensive experiments show that Compact Convolutional Networks (CCNs) neutralize multiple types of attacks, and perform better than existing methods in defending adversarial attacks, without incurring any additional training overhead compared to CNNs.Item Multi-modal Active Authentication of Smartphone Users(2018) Mahbub, Upal; Chellappa, Rama; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)With the increasing usage of smartphones not only as communication devices but also as the port of entry for a wide variety of user accounts at different information sensitivity levels, the need for hassle-free authentication is on the rise. Going beyond the traditional one-time authentication concept, active authentication (AA) schemes are emerging which authenticates users periodically in the background without the need for any user interaction. The purpose of this research is to explore different aspects of the AA problem and develop viable solutions by extracting unique biometric traits of the user from the wide variety of usage data obtained from Smartphone sensors. The key aspects of our research are the development of different components of user verification algorithms based on (a) face images from the front camera and (b) data from modalities other than the face. Since generic face detection algorithms do not perform very well in the mobile domain due to a significant presence of occluded and partially visible faces, we propose facial segment-based face detection technique to handle the challenge of partial faces in the mobile domain. We have developed three increasingly accurate proposal-based face detection methods, namely Facial Segment-based Face Detector (FSFD), SegFace and DeepSegFace, respectively, which perform binary classification on the results of a novel proposal generator that utilizes facial segments to obtain face-proposals. We also propose the Deep Regression-based User Image Detector (DRUID) network which shifts from the classification to the regression paradigm to avoid the need for proposal generation and thereby, achieves better processing speed and accuracy. DeepSegFace and DRUID have unique network architectures with customized loss functions and utilize a novel data augmentation scheme to train on a relatively small amount of data. The proposed methods, especially DRUID show superior performance over other state-of-the-art face detectors in terms of precision-recall and ROC curve on two mobile face datasets. We extended the concept of facial-segments to facial attribute detection for partially visible faces, a topic rarely addressed in the literature. We developed a deep convolutional neural network-based method named Segment-wise, Partial, Localized Inference in Training Facial Attribute Classification Ensembles (SPLITFACE) to detect attributes reliably from partially occluded faces. Taking several facial segments and the full face as input, SPLITFACE takes a data-driven approach to determine which attributes are localized in which facial segments. The unique architecture of the network allows each attribute to be predicted by multiple segments, which permits the implementation of committee machine techniques for combining local and global decisions to boost performance. Our evaluations on the full CelebA and LFWA datasets and their modified partial-visibility versions show that SPLITFACE significantly outperforms other recent attribute detection methods, especially for partial faces and for cross-domain experiments. We also explored the potentials of two less popular modalities namely, location history and application-usage, for active authentication. Aiming to discover the pattern of life of a user, we processed the location traces into separate state space models for each user and developed the Marginally Smoothed Hidden Markov Model (MSHMM) algorithm to authenticate the current user based on the most recent sequence of observations. The method takes into consideration the sparsity of the available data, the transition phases between states, the timing information and also the unforeseen states. We looked deeper into the impact of unforeseen and unknown states in another research work where we evaluated the feasibility of application usage behavior of the users as a potential solution to the active authentication problem. Our experiments show that it is essential to take unforeseen states into account when designing an authentication system with sparse data and marginal-smoothing techniques are very useful in this regard. We conclude this dissertation with the description of some ongoing efforts and future directions of research related the topics discussed in addition to a summary of all the contributions and impacts of this research work.Item Partial Face Detection and Illumination Estimation(2018) Sarkar, Sayantan; Chellappa, Rama; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Face Analysis has long been a crucial component of many security applications. In this work, we shall propose and explore some face analysis algorithms which are applicable to two different security problems, namely Active Authentication and Image Tampering Detection. In the first section, we propose two algorithms, “Deep Feature based Face Detection for Mobile Devices” and “DeepSegFace” that are useful in detecting partial faces such as those seem in typical Active Authentication scenarios. In the second section, we propose an algorithm to detect discrepancies in illumination conditions given two face images, and use that as an indication to decide if an image has been tampered by transplanting faces. We also extend the illumination detection algorithm by proposing an adversarial data augmentation scheme. We show the efficacy of the proposed algorithms by evaluating them on multiple datasets.