Search

Designed as extremely deep architectures, deep residual networks which provide a rich visual representation and offer robust convergence behaviors have recently achieved exceptional performance in numerous computer vision problems. Being directly applied to a scene labeling problem, however, they were limited to capture long-range contextual dependence, which is a critical aspect. To address this issue, we propose a novel approach, Contextual Recurrent Residual Networks (CRRN) which is able to simultaneously handle rich visual representation learning and long-range context modeling within a fully end-to-end deep network. Furthermore, our proposed end-to-end CRRN is completely trained from scratch, without using any pre-trained models in contrast to most existing methods usually fine-tuned from the state-of-the-art pre-trained models, e.g. VGG-16, ResNet, etc. The experiments are conducted on four challenging scene labeling datasets, i.e. SiftFlow, CamVid, Stanford background and SUN datasets, and compared against various state-of-the-art scene labeling methods. [ABSTRACT FROM AUTHOR]

This paper presents a Grammar-aware Driver Parsing (GDP) algorithm, with deep features, to provide a novel driver behavior situational awareness system (DB-SAW). A deep model is first trained to extract highly discriminative features of the driver. Then, a grammatical structure on the deep features is defined to be used as prior knowledge for a semi-supervised proposal candidate generation. The Region with Convolutional Neural Networks (R-CNN) method is ultimately utilized to precisely segment parts of the driver. The proposed method not only aims to automatically find parts of the driver in challenging “drivers in the wild” databases, i.e. the standardized Strategic Highway Research Program (SHRP-2) and the challenging Vision for Intelligent Vehicles and Application (VIVA), but is also able to investigate seat belt usage and the position of the driver's hands (on a phone vs on a steering wheel). We conduct experiments on various applications and compare our GDP method against other state-of-the-art detection and segmentation approaches, i.e. SDS [1] , CRF-RNN [2] , DJTL [3] , and R-CNN [4] on SHRP-2 and VIVA databases. [ABSTRACT FROM AUTHOR]

This paper presents a robust, fully automatic and semi self-training system to detect and segment facial beard/moustache simultaneously in challenging facial images. Based on the observation that some certain facial areas, e.g. cheeks, do not typically contain any facial hair whereas the others, e.g. brows, often contain facial hair, a self-trained model is first built using a testing image itself. To overcome the limitation of that facial hairs in brows regions and beard/moustache regions are different in length, density, color, etc., a pre-trained model is also constructed using training data. The pre-trained model is only pursued when the self-trained model produces low confident classification results. In the proposed system, we employ the superpixel together a combination of two classifiers, i.e. Random Ferns (rFerns) and Support Vector Machines (SVM) to obtain good classification performance as well as improve time efficiency. A feature vector, consisting of Histogram of Gabor (HoG) and Histogram of Oriented Gradient of Gabor (HOGG) at different directions and frequencies, is generated from both the bounding box of the superpixel and the super pixel foreground. The segmentation result is then refined by our proposed aggregately searching strategy in order to deal with inaccurate landmarking points. Experimental results have demonstrated the robustness and effectiveness of the proposed system. It is evaluated in images drawn from three entire databases i.e. the Multiple Biometric Grand Challenge (MBGC) still face database, the NIST color Facial Recognition Technology FERET database and a large subset from Pinellas County database. [ABSTRACT FROM AUTHOR]

In this paper, we propose a novel joint formulation of feature-based active contour (FAC) and prior shape constraints (CS) for lips/mouth segmentation in the wild. Our proposed SC-FAC model is able to robustly segment the lips/mouth that belongs to a given mouth shape space while minimizing the energy functional. The shape space is defined by a 2D Modified Active Shape Model (MASM) whereas the active contour model is based on the Chan-Vese functional. Our SC-FAC energy functional is able to overcome the drawback of noise while minimizing the fitting forces under the shape constraints. We conducted our experiments on images captured under challenging conditions such as varying illumination, low contrast, facial expression, low resolution, blurring, wearing beard/moustache and cosmetic affection from the MBGC, VidTIMIT, JAFFE, and LFW databases. The results from our studies indicate that the proposed SC-FAC model is reliable and accurately perform prior shape weak object segmentation. The average performance of the mouth segmentation using proposed SC-FAC on 1918 images from the MBGC database under different illuminations, expressions, and complex background reaches to a Precision of 91.30%, a Recall of 91.32% and an F-measure of 90.62%. [ABSTRACT FROM AUTHOR]

A reliable and accurate biometric identification system must be able to distinguish individuals even in situations where their biometric signatures are very similar. However, the strong similarity in the facial appearance of twins has complicated facial feature based recognition and has even compromised commercial face recognition systems. This paper addresses the above problem and proposes two novel methods to distinguish identical twins using (1) facial aging features and (2) asymmetry decomposition features. Facial aging features are extracted using Gabor filters from regions of the face that typically exhibit wrinkles and laugh lines, while Facial asymmetry decomposition based features are obtained by projecting the difference between the two left sides (consisting of the left half of the face and its mirror) and two right sides (consisting of the right half of the face and its mirror) of a face onto a subspace. Feature vectors obtained using these methods were used for classification. Experiments conducted on images of five types of twins from the University of Notre Dame ND-Twins database indicate that both our proposed approaches achieve high identification rates and are hence quite promising at distinguishing twins. [ABSTRACT FROM AUTHOR]

Abstract: Steganography is the art and science of hiding data into information. The secret message is hidden in such a way that no one can apart from the sender or the intended recipient. The least significant bit (LSB) substitution mechanism is the most common steganographic technique for embedding a secret message in an image with high capacity, while the human visual system (HVS) would be unable to notice the hidden message in the cover image. In this paper, besides employing the LSB substitution technique as a fundamental stage, we take advantage of edge detection technique. The experimental results show that the proposed scheme not only achieves high embedding capacity but also enhances the quality of the stego image from the HVS by an edge detection technique. Moreover, based on that the secret message is replaced with different LSBs, our scheme can effectively resist the image steganalysis. [Copyright &y& Elsevier]

Abstract: In this paper, we propose a novel verifiable secret sharing (VSS) scheme, which not only protects a secret image but also allows users to verify the restored secret image in the revealing and verifying phase, for all binary, grayscale and color images. Error diffusion and image clustering techniques are adopted to achieve our objective. Experimental results and discussions show that the proposed scheme, with its smaller shadow size and lower computational complexity, obviously outperforms previous VSS schemes designed either with or without the cheating prevention mechanism. Moreover, the use of a halftone logo gives an efficient solution to verifying whether the restored secret image is correct by using a halftone logo. [Copyright &y& Elsevier]

This paper presents a novel subject-dependent deep aging path (SDAP), which inherits the merits of both generative probabilistic modeling and inverse reinforcement learning to model the facial structures and the longitudinal face aging process of a given subject. The proposed SDAP is optimized using tractable log-likelihood objective functions with convolutional neural networks (CNNs) based deep feature extraction. Instead of applying a fixed aging development path for all input faces and subjects, SDAP is able to provide the most appropriate aging development path for individual subject that optimizes the reward aging formulation. Unlike previous methods that can take only one image as the input, SDAP further allows multiple images as inputs, i.e. all information of a subject at either the same or different ages, to produce the optimal aging path for the given subject. Finally, SDAP allows efficiently synthesizing in-the-wild aging faces. The proposed model is experimented in both tasks of face aging synthesis and cross-age face verification. The experimental results consistently show SDAP achieves the state-of-the-art performance on numerous face aging databases, i.e. FG-NET, MORPH, aging faces in the wild (AGFW), and cross-age celebrity dataset (CACD). Furthermore, we also evaluate the performance of SDAP on large-scale Megaface challenge to demonstrate the advantages of the proposed solution. [ABSTRACT FROM AUTHOR]