RESUMO
This paper proposes a 3D face alignment of 2D face images in the wild with noisy landmarks. The objective is to recognize individuals from their single profile image. We first proceed by extracting more than 68 landmarks using a bag of features. This allows us to obtain a bag of visible and invisible facial keypoints. Then, we reconstruct a 3D face model and get a triangular mesh by meshing the obtained keypoints. For each face, the number of keypoints is not the same, which makes this step very challenging. Later, we process the 3D face using butterfly and BPA algorithms to make correlation and regularity between 3D face regions. Indeed, 2D-to-3D annotations give much higher quality to the 3D reconstructed face model without the need for any additional 3D Morphable models. Finally, we carry out alignment and pose correction steps to get frontal pose by fitting the rendered 3D reconstructed face to 2D face and performing pose normalization to achieve good rates in face recognition. The recognition step is based on deep learning and it is performed using DCNNs, which are very powerful and modern, for feature learning and face identification. To verify the proposed method, three popular benchmarks, YTF, LFW, and BIWI databases, are tested. Compared to the best recognition results reported on these benchmarks, our proposed method achieves comparable or even better recognition performances.
RESUMO
Appropriate modeling of a surveillance scene is essential for the detection of anomalies in road traffic. Learning usual paths can provide valuable insight into road traffic conditions and thus can help in identifying unusual routes taken by commuters/vehicles. If usual traffic paths are learned in a nonparametric way, manual interventions in road marking can be avoided. In this paper, we propose an unsupervised and nonparametric method to learn the frequently used paths from the tracks of moving objects in Θ(kn) time, where k denotes the number of paths and n represents the number of tracks. In the proposed method, temporal dependencies of the moving objects are considered to make the clustering meaningful using temporally incremental gravity model (TIGM). In addition, the distance-based scene learning makes it intuitive to estimate the model parameters. Further, we have extended the TIGM hierarchically as a dynamically evolving model (DEM) to represent notable traffic dynamics of a scene. The experimental validation reveals that the proposed method can learn a scene quickly without prior knowledge about the number of paths ( k ). We have compared the results with various state-of-the-art methods. We have also highlighted the advantages of the proposed method over the existing techniques popularly used for designing traffic monitoring applications. It can be used for administrative decision making to control traffic at junctions or crowded places and generate alarm signals, if necessary.
RESUMO
Stochastic gradient descent (SGD) is one of the core techniques behind the success of deep neural networks. The gradient provides information on the direction in which a function has the steepest rate of change. The main problem with basic SGD is to change by equal-sized steps for all parameters, irrespective of the gradient behavior. Hence, an efficient way of deep network optimization is to have adaptive step sizes for each parameter. Recently, several attempts have been made to improve gradient descent methods such as AdaGrad, AdaDelta, RMSProp, and adaptive moment estimation (Adam). These methods rely on the square roots of exponential moving averages of squared past gradients. Thus, these methods do not take advantage of local change in gradients. In this article, a novel optimizer is proposed based on the difference between the present and the immediate past gradient (i.e., diffGrad). In the proposed diffGrad optimization technique, the step size is adjusted for each parameter in such a way that it should have a larger step size for faster gradient changing parameters and a lower step size for lower gradient changing parameters. The convergence analysis is done using the regret bound approach of the online learning framework. In this article, thorough analysis is made over three synthetic complex nonconvex functions. The image categorization experiments are also conducted over the CIFAR10 and CIFAR100 data sets to observe the performance of diffGrad with respect to the state-of-the-art optimizers such as SGDM, AdaGrad, AdaDelta, RMSProp, AMSGrad, and Adam. The residual unit (ResNet)-based convolutional neural network (CNN) architecture is used in the experiments. The experiments show that diffGrad outperforms other optimizers. Also, we show that diffGrad performs uniformly well for training CNN using different activation functions. The source code is made publicly available at https://github.com/shivram1987/diffGrad.
