Search

A simplified model of heat and mass transfer of integrated solar still with an underground heat exchanger (GHEX) is presented. The solar still has 70 m length and 0.5 m2 inlet area. Moreover, the underground heat exchanger was 70 m length and 0.2 m diameter. The model includes a theoretical analysis in the transient mode for both the solar still and underground heat exchanger under warm climate region throughout the year. This system uses solar energy to saturate the air in the solar still, which is conducted into a buried pipe. Fresh water will be formed by condensation as long as the surface temperature of the pipe wall is less or equal to the dew point temperature of the humid air. A numerical method has been employed to solve the governing equations of heat and mass transfer characteristics in both the solar still and the underground heat exchanger. The soil temperature was taken as a function of soil depth and time along the year. The amount of fresh water produced was found strongly depends on solar intensity. It was respectively 4.7 (kg/m day) and 0.32 (kg/m day) for the highest and lowest values of solar radiation during the year. A good agreement has been shown when the results were compared with previous work.

A comparison between artificial neural network (ANN) and multiple linear regression (MLR) models was employed to predict the heat of combustion, and the gross and net heat values, of a diesel fuel engine, based on the chemical composition of the diesel fuel. One hundred and fifty samples of Iraqi diesel provided data from chromatographic analysis. Eight parameters were applied as inputs in order to predict the gross and net heat combustion of the diesel fuel. A trial-and-error method was used to determine the shape of the individual ANN. The results showed that the prediction accuracy of the ANN model was greater than that of the MLR model in predicting the gross heat value. The best neural network for predicting the gross heating value was a back-propagation network (8-8-1), using the Levenberg�Marquardt algorithm for the second step of network training. R = 0.98502 for the test data. In the same way, the best neural network for predicting the net heating value was a back-propagation network (8-5-1), using the Levenberg�Marquardt algorithm for the second step of network training. R = 0.95112 for the test data.

Zero-shot learning aims to perform visual reasoning about unseen objects. In-line with the success of deep learning on object recognition problems, several end-to-end deep models for zero-shot recognition have been proposed in the literature. These models are successful in predicting a single unseen label given an input image but do not scale to cases where multiple unseen objects are present. Here, we focus on the challenging problem of zero-shot image tagging, where multiple labels are assigned to an image, that may relate to objects, attributes, actions, events, and scene type. Discovery of these scene concepts requires the ability to process multi-scale information. To encompass global as well as local image details, we propose an automatic approach to locate relevant image patches and model image tagging within the Multiple Instance Learning (MIL) framework. To the best of our knowledge, we propose the first end-to-end trainable deep MIL framework for the multi-label zero-shot tagging problem. We explore several alternatives for instance-level evidence aggregation and perform an extensive ablation study to identify the optimal pooling strategy. Due to its novel design, the proposed framework has several interesting features: 1) unlike previous deep MIL models, it does not use any off-line procedure (e.g., Selective Search or EdgeBoxes) for bag generation. 2) During test time, it can process any number of unseen labels given their semantic embedding vectors. 3) Using only image-level seen labels as weak annotation, it can produce a localized bounding box for each predicted label. We experiment with the large-scale NUS-WIDE and MS-COCO datasets and achieve superior performance across conventional, zero-shot, and generalized zero-shot tagging tasks.

Convolutional Neural Networks have achieved significant success across multiple computer vision tasks. However, they are vulnerable to carefully crafted, human-imperceptible adversarial noise patterns which constrain their deployment in critical security-sensitive systems. This paper proposes a computationally efficient image enhancement approach that provides a strong defense mechanism to effectively mitigate the effect of such adversarial perturbations. We show that deep image restoration networks learn mapping functions that can bring off-the-manifold adversarial samples onto the natural image manifold, thus restoring classification towards correct classes. A distinguishing feature of our approach is that, in addition to providing robustness against attacks, it simultaneously enhances image quality and retains models performance on clean images. Furthermore, the proposed method does not modify the classifier or requires a separate mechanism to detect adversarial images. The effectiveness of the scheme has been demonstrated through extensive experiments, where it has proven a strong defense in gray-box settings. The proposed scheme is simple and has the following advantages: (1) it does not require any model training or parameter optimization, (2) it complements other existing defense mechanisms, (3) it is agnostic to the attacked model and attack type and (4) it provides superior performance across all popular attack algorithms. Our codes are publicly available at https://github.com/aamir-mustafa/super-resolution-adversarial-defense., Published in IEEE Transactions in Image Processing

COVID-19 vaccine hesitancy is a significant threat to the efforts that have been taken to combat the pandemic. This study assessed the acceptability of the COVID-19 vaccine among university students in Saudi Arabia. A cross-sectional online survey using a Google Form was conducted between 2 April and 23 April 2021. A snowball technique was used to recruit participants for this study. The final sample consisted of 407 participants. More than one-third of the participants (36.1%) had received the COVID-19 vaccine, and 13.3% had registered to receive the vaccine. Of the participants who were not yet vaccinated (n = 260), 90.4% indicated that they would like to be vaccinated when given the opportunity. Of the unvaccinated participants, 82.3% trusted the COVID-19 vaccines that had been provided in Saudi Arabia. The most reported reasons for the participants’ acceptance to receive the COVID-19 vaccine included preventive purposes (95.8%), a belief in the safety of the vaccines (84.3%), and the availability of public awareness information regarding the vaccines (77.3%). A small portion of participants (6.1%) were refusing to receive the vaccine due to the potential long-term side effects (92.0%) and expedited vaccine trials (80.0%). Acceptability of the COVID-19 vaccine was strongly associated with participants who regularly received the flu vaccine (p &lt, 0.05). All other demographic variables were not statistically associated with the acceptability of the COVID-19 vaccine. In conclusion, it would be appropriate for universities to launch peer programs to urge reluctant students to receive the vaccine voluntarily. In terms of further research, it is valuable to follow up with unvaccinated participants to investigate if they received the vaccine since the data were collected, and their reasons for doing so. This research would reveal changes toward vaccine acceptability over time and any related determinants. Future research should consider students from non-Arabic speaking backgrounds.

OBJECTIVE: To conduct a multicentre study to establish the critical concentration (CC) for clofazimine (CFZ) for drug susceptibility testing (DST) of Mycobacterium tuberculosis on the MGIT (TM) 960 (TM) system using the distribution of minimum inhibitory concentrations (MIC) and genotypic analyses of Rv0678 mutations. DESIGN: In phase I of the study, the MIC distribution of laboratory strains (H37Rv and in vitro-selected Rv0678 mutants) and clinical pan-susceptible isolates were determined (n = 70). In phase II, a tentative CC for CFZ (n = 55) was proposed. In phase III, the proposed CC was validated using clinical drug-resistant tuberculosis (DR-TB) isolates stratified by Rv0678 mutation (n = 85). RESULTS AND CONCLUSION: The MIC distribution of CFZ for laboratory and clinical pan-susceptible strains ranged between 0.125 mu g/ml and 0.5 mu g/ml. As the MIC values of DR-TB isolates used for phase II ranged between 0.25 mu g/ml and 1 mu g/ml, a CC of 1 mu g/ml was proposed. Validation of the CC in phase III showed that probably susceptible and probably resistant Rv0678 mutants overlapped at 1 mu g/ml. We therefore recommend a CC of 1 mu g/ml, with additional testing at 0.5 mu g/ml to define an intermediate category. This was the first comprehensive study to establish a CC for routine phenotypic DST of CFZ using the MGIT960 system to guide therapeutic decisions.

Today’s, automated process planning is becoming a widely prevalent technology in modern manufacturing systems. Feature concept is a core required to realize a fully integrated CAD/CAM system. The contents of a feature’s information can simply be transferred from one manufacturing activity to another. in the last few years, Automated process planning based on features technology has become popular with its capacity to combine design and manufacturing systematically. Despite its advantages, it is still difficult to recognize and extract manufacturing features automatically from the CAD system database. This paper addressed the development of a Feature-Based Process Planning System (FBPPS) integrated with CAD systems Automatically. The developed system used the knowledge-base method to execute process planning tasks that include; process selection, process sequences, machine tool selection, tool selection, and cutting conditions. Feature recognize module based on syntactic recognition technique has been introduced using CAD database of 3D component geometry model. The process planning module has been implemented using decision-making procedures based on the feature recognition stage. The proposed framework was implemented on personal computers using AutoCAD 2014 as the CAD modeling system and Visual-Basic 2010 as the programming environment to build the decision-making knowledge base. A 3D CAD model of the prismatic mechanical part has been selected as a case study to examine and evaluate the FBPPS ability. The final results of the test show the success of the proposed system on manufacturing feature recognition and generate a process plan automatically.

The existing zero-shot detection approaches project visual features to the semantic domain for seen objects, hoping to map unseen objects to their corresponding semantics during inference. However, since the unseen objects are never visualized during training, the detection model is skewed towards seen content, thereby labeling unseen as background or a seen class. In this work, we propose to synthesize visual features for unseen classes, so that the model learns both seen and unseen objects in the visual domain. Consequently, the major challenge becomes, how to accurately synthesize unseen objects merely using their class semantics? Towards this ambitious goal, we propose a novel generative model that uses class-semantics to not only generate the features but also to discriminatively separate them. Further, using a unified model, we ensure the synthesized features have high diversity that represents the intra-class differences and variable localization precision in the detected bounding boxes. We test our approach on three object detection benchmarks, PASCAL VOC, MSCOCO, and ILSVRC detection, under both conventional and generalized settings, showing impressive gains over the state-of-the-art methods. Our codes are available at https://github.com/nasir6/zero_shot_detection.

Previous work on novel object detection considers zero or few-shot settings where none or few examples of each category are available for training. In real world scenarios, it is less practical to expect that ‘all’ the novel classes are either unseen or have few-examples. Here, we propose a more realistic setting termed ‘Any-shot detection’, where totally unseen and few-shot categories can simultaneously co-occur during inference. Any-shot detection offers unique challenges compared to conventional novel object detection such as, a high imbalance between unseen, few-shot and seen object classes, susceptibility to forget base-training while learning novel classes and distinguishing novel classes from the background. To address these challenges, we propose a unified any-shot detection model, that can concurrently learn to detect both zero-shot and few-shot object classes. Our core idea is to use class semantics as prototypes for object detection, a formulation that naturally minimizes knowledge forgetting and mitigates the class-imbalance in the label space. Besides, we propose a rebalanced loss function that emphasizes difficult few-shot cases but avoids overfitting on the novel classes to allow detection of totally unseen classes. Without bells and whistles, our framework can also be used solely for Zero-shot object detection and Few-shot object detection tasks. We report extensive experiments on Pascal VOC and MS-COCO datasets where our approach is shown to provide significant improvements.

Generic object counting in natural scenes is a challenging computer vision problem. Existing approaches either rely on instance-level supervision or absolute count information to train a generic object counter. We introduce a partially supervised setting that significantly reduces the supervision level required for generic object counting. We propose two novel frameworks, named lower-count (LC) and reduced lower-count (RLC), to enable object counting under this setting. Our frameworks are built on a novel dual-branch architecture that has an image classification and a density branch. Our LC framework reduces the annotation cost due to multiple instances in an image by using only lower-count supervision for all object categories. Our RLC framework further reduces the annotation cost arising from large numbers of object categories in a dataset by only using lower-count supervision for a subset of categories and class-labels for the remaining ones. The RLC framework extends our dual-branch LC framework with a novel weight modulation layer and a category-independent density map prediction. Experiments are performed on COCO, Visual Genome and PASCAL 2007 datasets. Our frameworks perform on par with state-of-the-art approaches using higher levels of supervision. Additionally, we demonstrate the applicability of our LC supervised density map for image-level supervised instance segmentation., Comment: Accepted by IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2020. First two authors contributed equally. Corresponding author: Guolei Sun. This work is a journal extension of our CVPR 2019 paper arXiv:1903.02494

3D object instance reconstruction from a cluttered 2D scene image is an ill-posed problem. The main challenge is posed by the lack of geometric information in color images and heavy occlusions that lead to incomplete shape details. To deal with this problem, existing works on 3D instance reconstruction directly learn the mapping between the intensity image and the corresponding 3D volume model. Different from these works, we propose to explicitly incorporate 2.5D geometric cues, such as the surface normal, relative depth, and height, while generating full 3D shapes from 2D images. With an intermediate step focused on estimating these 2.5D geometric features, we propose a novel convolutional neural network design that progressively moves from 2D to full 3D estimation. Our model automatically generates instance-specific surface normal maps, relative depth, and height that are compactly encoded within our network design and consequently used to improve the 3D instance reconstruction. Our experimental results on the large-scale synthetic SUNCG dataset and the real-world NYU depth v2 dataset demonstrate the effectiveness of the proposed approach where it beats the state-of-the-art Factored3D network [15].

Adversarial examples can cause catastrophic mistakes in Deep Neural Network (DNNs) based vision systems e.g., for classification, segmentation and object detection. The vulnerability of DNNs against such attacks can prove a major roadblock towards their real-world deployment. Transferability of adversarial examples demand generalizable defenses that can provide cross-task protection. Adversarial training that enhances robustness by modifying target model's parameters lacks such generalizability. On the other hand, different input processing based defenses fall short in the face of continuously evolving attacks. In this paper, we take the first step to combine the benefits of both approaches and propose a self-supervised adversarial training mechanism in the input space. By design, our defense is a generalizable approach and provides significant robustness against the \textbf{unseen} adversarial attacks (\eg by reducing the success rate of translation-invariant \textbf{ensemble} attack from 82.6\% to 31.9\% in comparison to previous state-of-the-art). It can be deployed as a plug-and-play solution to protect a variety of vision systems, as we demonstrate for the case of classification, segmentation and detection. Code is available at: {\small\url{https://github.com/Muzammal-Naseer/NRP}}., Comment: CVPR-2020 (Oral). Code this http https://github.com/Muzammal-Naseer/NRP}

Humans can continuously learn new knowledge as their experience grows. In contrast, previous learning in deep neural networks can quickly fade out when they are trained on a new task. In this paper, we hypothesize this problem can be avoided by learning a set of generalized parameters, that are neither specific to old nor new tasks. In this pursuit, we introduce a novel meta-learning approach that seeks to maintain an equilibrium between all the encountered tasks. This is ensured by a new meta-update rule which avoids catastrophic forgetting. In comparison to previous meta-learning techniques, our approach is task-agnostic. When presented with a continuum of data, our model automatically identifies the task and quickly adapts to it with just a single update. We perform extensive experiments on five datasets in a class-incremental setting, leading to significant improvements over the state of the art methods (e.g., a 21.3% boost on CIFAR100 with 10 incremental tasks). Specifically, on large-scale datasets that generally prove difficult cases for incremental learning, our approach delivers absolute gains as high as 19.1% and 7.4% on ImageNet and MS-Celeb datasets, respectively., Accepted to CVPR 2020

Deep neural networks can easily be fooled by an adversary with minuscule perturbations added to an input image. The existing defense techniques suffer greatly under white-box attack settings, where an adversary has full knowledge of the network and can iterate several times to find strong perturbations. We observe that the main reason for the existence of such vulnerabilities is the close proximity of different class samples in the learned feature space of deep models. This allows the model decisions to be completely changed by adding an imperceptible perturbation to the inputs. To counter this, we propose to class-wise disentangle the intermediate feature representations of deep networks, specifically forcing the features for each class to lie inside a convex polytope that is maximally separated from the polytopes of other classes. In this manner, the network is forced to learn distinct and distant decision regions for each class. We observe that this simple constraint on the features greatly enhances the robustness of learned models, even against the strongest white-box attacks, without degrading the classification performance on clean images. We report extensive evaluations in both black-box and white-box attack scenarios and show significant gains in comparison to state-of-the-art defenses.

The present study was designed to explore the neuroprotective properties of Aconitum napellus (Ranunculaceae). The plant detoxification was done using either water, or cow or goat milk as per the Ayurvedic shodhana method. The evaluation of the neuroprotective role of A. napellus was performed on diabetic neuropathy induced by streptozotocin in Sprague Dawley (SD) rats. Body mass, blood sugar level, oral glucose tolerance test, hyperalgesia, cold allodynia, motor co-ordination test, and locomotor activity, oxidative biomarkers (TBARS, reduced glutathione, catalase and superoxide dismutase) and sciatic nerve histomorphology were assessed. The in vitro studies were done on human neuroblastoma cell line SHSY-5Y and used an MTT assay to assess the antiproliferative activity of different extracts. Results suggest that the goat milk treated chloroform extract has less percentage of aconitine. After administration of the detoxified chloroform extract to the diabetic animals, there was a significant improvement in the myelination and degenerative changes of the nerve fibers along with behavioral changes (p &lt, 0.05 as compared with diabetic control group). The findings of the in vitro research show an effective neuroprotective role of A. napellus. This suggests that A. napellus should be further investigated for its effect in diabetic pathology.

Cardiovascular disorders are among the most important causes of sudden death and adult disability worldwide. Abdominal aortic aneurysm (AAA) is a critical clinical condition where the aorta dilates beyond 50% of its normal diameter and leads to a risk of rupture. In this study, we performed fluid-structure interaction (FSI) analysis on an eccentric computational AAA model in order to investigate the effects of wall thickness on AAA wall stresses, which are critically important to estimate the rupture risk. For this purpose, we modeled the problem domain using finite element analysis, and coupled the solutions of fluid and structure domains for improving the accuracy of results. ANSYS commercial finite element analysis software was used for modeling, solving, and post-processing the results. Expanded diameter in AAA sac resulted in altered hemodynamics. Wall shear stresses (WSS) caused by the flow are quite low on the AAA sac, which may deteriorate the endothelial cell regeneration and vascular remodeling in the long term. It is concluded that the most critical region for the rupture risk is the posterior distal end of AAA sac due to being exposed to peak mechanical stresses during the cardiac cycle. Obtained results shed light in understanding the rupture risk assessment of AAA. Qatar University Research Fund, Qatar University

The availability of large-scale datasets has helped unleash the true potential of deep convolutional neural networks (CNNs). However, for the single-image denoising problem, capturing a real dataset is an unacceptably expensive and cumbersome procedure. Consequently, image denoising algorithms are mostly developed and evaluated on synthetic data that is usually generated with a widespread assumption of additive white Gaussian noise (AWGN). While the CNNs achieve impressive results on these synthetic datasets, they do not perform well when applied on real camera images, as reported in recent benchmark datasets. This is mainly because the AWGN is not adequate for modeling the real camera noise which is signal-dependent and heavily transformed by the camera imaging pipeline. In this paper, we present a framework that models camera imaging pipeline in forward and reverse directions. It allows us to produce any number of realistic image pairs for denoising both in RAW and sRGB spaces. By training a new image denoising network on realistic synthetic data, we achieve the state-of-the-art performance on real camera benchmark datasets. The parameters in our model are ~5 times lesser than the previous best method for RAW denoising. Furthermore, we demonstrate that the proposed framework generalizes beyond image denoising problem e.g., for color matching in stereoscopic cinema. The source code and pre-trained models are available at https://github.com/swz30/CycleISP., Comment: CVPR 2020 (Oral)

Class imbalance is a common problem in the case of real-world object detection and classification tasks. Data of some classes are abundant, making them an overrepresented majority, and data of other classes are scarce, making them an underrepresented minority. This imbalance makes it challenging for a classifier to appropriately learn the discriminating boundaries of the majority and minority classes. In this paper, we propose a cost-sensitive (CoSen) deep neural network, which can automatically learn robust feature representations for both the majority and minority classes. During training, our learning procedure jointly optimizes the class-dependent costs and the neural network parameters. The proposed approach is applicable to both binary and multiclass problems without any modification. Moreover, as opposed to data-level approaches, we do not alter the original data distribution, which results in a lower computational cost during the training process. We report the results of our experiments on six major image classification data sets and show that the proposed approach significantly outperforms the baseline algorithms. Comparisons with popular data sampling techniques and CoSen classifiers demonstrate the superior performance of our proposed method.

Background: Spindle epithelial tumor with thymus-like differentiation (SETTLE) is a very rare neoplasm of the thyroid first described by Chan and Rosai in 1991. Renal involvement usually occurs as part of a disseminated metastatic picture. SETTLE is believed to arise from branchial pouch or thymic remnants and shows primitive thymic differentiation. Presentation: A 41-year-old male presented with a thyroid mass in 2005. The patient underwent a total thyroidectomy followed by radiation therapy. Pathology concluded papillary thyroid cancer. In 2011, the patient was admitted with worsening cough. CT showed lung involvement, multiple mediastinal lymph nodes, and a left renal mass. A biopsy established the diagnosis of SETTLE. Four cycles of platinum chemotherapy were given. In 2014, the patient was referred to Urology with progressive flank pain and hematuria. CT showed a progression of the renal mass, and results of an U/S guided biopsy of the kidney were consistent with SETTLE. An elective radical nephrectomy was performed in late 2016 to relieve symptoms. The patient is currently being followed as an outpatient. Conclusion: Due to the rare nature of SETTLE, algorithms for the diagnosis, treatment, and outcomes are not established. Although SETTLE presents as a low-grade malignancy, this case, as well as other reported cases with a long follow-up duration, suggested that it can metastasize many years after initial diagnosis. More data is needed to elucidate the risk and prognosis of renal metastasis in SETTLE.

In dipole-antenna array signal-processing algorithm development, research has focused on “short dipoles” whose physical length ( $L$ ) is under (1/10) of a wavelength $\lambda $ . Such electrically “short” dipoles have very small input impedances, rendering such “short” dipoles to be inefficient radiators. Practical dipoles, with an electrical length of $({L}/{\lambda }) \in [{0.1, 1}]$ , have notably larger input impedance, hence making them better radiators. Of such practical dipoles, this paper investigates their use for azimuth-polar direction finding and for polarization estimation. This paper will first present the measurement model (i.e., array manifold) of a triad of such practical dipoles, collocated in space and orthogonally oriented. This paper will then develop the corresponding closed-form algorithms to estimate the bivariate azimuth–elevation direction-of-arrival or the bivariate polarization. Such closed-form algorithms previously have been unavailable in the existing literature for such a triad of electrically “long” dipoles of pragmatic radiation efficiency.

Land cover change monitoring is an important task from the perspective of regional resource monitoring, disaster management, land development, and environmental planning. In this paper, we analyze imagery data from remote sensing satellites to detect forest cover changes over a period of 29 years (1987–2015). Since the original data are severely incomplete and contaminated with artifacts, we first devise a spatiotemporal inpainting mechanism to recover the missing surface reflectance information. The spatial filling process makes use of the available data of the nearby temporal instances followed by a sparse encoding-based reconstruction. We formulate the change detection task as a region classification problem. We build a multiresolution profile (MRP) of the target area and generate a candidate set of bounding-box proposals that enclose potential change regions. In contrast to existing methods that use handcrafted features, we automatically learn region representations using a deep neural network in a data-driven fashion. Based on these highly discriminative representations, we determine forest changes and predict their onset and offset timings by labeling the candidate set of proposals. Our approach achieves the state-of-the-art average patch classification rate of 91.6% (an improvement of ~16%) and the mean onset/offset prediction error of 4.9 months (an error reduction of five months) compared with a strong baseline. We also qualitatively analyze the detected changes in the unlabeled image regions, which demonstrate that the proposed forest change detection approach is scalable to new regions.

Face recognition from image sets has numerous real-life applications including recognition from security and surveillance systems, multi-view camera networks and personal albums. An image set is an unordered collection of images (e.g., video frames, images acquired over long term observations and personal albums) which exhibits a wide range of appearance variations. The main focus of the previously developed methods has therefore been to find a suitable representation to optimally model these variations. This paper argues that such a representation could not necessarily encode all of the information contained in the set. The paper, therefore, suggests a different approach which does not resort to a single representation of an image set. Instead, the images of the set are retained in their original form and an efficient classification strategy is developed which extends well-known simple binary classifiers for the task of multi-class image set classification. Unlike existing binary to multi-class extension strategies, which require multiple binary classifiers to be trained over a large number of images, the proposed approach is efficient since it trains only few binary classifiers on very few images. Extensive experiments and comparisons with existing methods show that the proposed approach achieves state of the art performance for image set classification based face and object recognition on a number of challenging datasets.

The main requisite for fine-grained recognition task is to focus on subtle discriminative details that make the subordinate classes different from each other. We note that existing methods implicitly address this requirement and leave it to a data-driven pipeline to figure out what makes a subordinate class different from the others. This results in two major limitations: First, the network focuses on the most obvious distinctions between classes and overlooks more subtle inter-class variations. Second, the chance of misclassifying a given sample in any of the negative classes is considered equal, while in fact, confusions generally occur among only the most similar classes. Here, we propose to explicitly force the network to find the subtle differences among closely related classes. In this pursuit, we introduce two key novelties that can be easily plugged into existing end-to-end deep learning pipelines. On one hand, we introduce diversification block which masks the most salient features for an input to force the network to use more subtle cues for its correct classification. Concurrently, we introduce a gradient-boosting loss function that focuses only on the confusing classes for each sample and therefore moves swiftly along the direction on the loss surface that seeks to resolve these ambiguities. The synergy between these two blocks helps the network to learn more effective feature representations. Comprehensive experiments are performed on five challenging datasets. Our approach outperforms existing methods using similar experimental setting on all five datasets., To appear in AAAI 2020

© 1999-2012 IEEE. Vision-based person re-identification aims to match a person's identity across multiple images, which is a fundamental task in multimedia content analysis and retrieval. Deep neural networks have recently manifested great potential in this task. However, a major bottleneck of existing supervised deep networks is their reliance on a large amount of annotated training data. Manual labeling for person identities in large-scale surveillance camera systems is quite challenging and incurs significant costs. Some recent studies adopt generative model outputs as training data augmentation. To more effectively use these synthetic data for an improved feature learning and re-identification performance, this paper proposes a novel feature affinity-based pseudo labeling method with two possible label encodings. To the best of our knowledge, this is the first study that employs pseudo-labeling by measuring the affinity of unlabeled samples with the underlying clusters of labeled data samples using the intermediate feature representations from deep networks. We propose training the network with the joint supervision of cross-entropy loss together with a center regularization term, which not only ensures discriminative feature representation learning but also simultaneously predicts pseudo-labels for unlabeled data. We show that both label encodings can be learned in a unified manner and help improve the overall performance. Our extensive experiments on three person re-identification datasets: Market-1501, DukeMTMC-reID, and CUHK03, demonstrate significant performance boost over the state-of-the-art person re-identification approaches.

Existing networks directly learn feature representations on 3D point clouds for shape analysis. We argue that 3D point clouds are highly redundant and hold irregular (permutation-invariant) structure, which makes it difficult to achieve inter-class discrimination efficiently. In this paper, we propose a two-faceted solution to this problem that is seamlessly integrated in a single `Blended Convolution and Synthesis' layer. This fully differentiable layer performs two critical tasks in succession. In the first step, it projects the input 3D point clouds into a latent 3D space to synthesize a highly compact and more inter-class discriminative point cloud representation. Since, 3D point clouds do not follow a Euclidean topology, standard 2/3D Convolutional Neural Networks offer limited representation capability. Therefore, in the second step, it uses a novel 3D convolution operator functioning inside the unit ball ($\mathbb{B}^3$) to extract useful volumetric features. We extensively derive formulae to achieve both translation and rotation of our novel convolution kernels. Finally, using the proposed techniques we present an extremely light-weight, end-to-end architecture that achieves compelling results on 3D shape recognition and retrieval., 10 pages: corrected typos and added affiliations. The IEEE Winter Conference on Applications of Computer Vision. 2020

Visual Question Answering (VQA) models employ attention mechanisms to discover image locations that are most relevant for answering a specific question. For this purpose, several multimodal fusion strategies have been proposed, ranging from relatively simple operations (e.g., linear sum) to more complex ones (e.g., Block). The resulting multimodal representations define an intermediate feature space for capturing the interplay between visual and semantic features, that is helpful in selectively focusing on image content. In this paper, we propose a question-agnostic attention mechanism that is complementary to the existing question-dependent attention mechanisms. Our proposed model parses object instances to obtain an `object map' and applies this map on the visual features to generate Question-Agnostic Attention (QAA) features. In contrast to question-dependent attention approaches that are learned end-to-end, the proposed QAA does not involve question-specific training, and can be easily included in almost any existing VQA model as a generic light-weight pre-processing step, thereby adding minimal computation overhead for training. Further, when used in complement with the question-dependent attention, the QAA allows the model to focus on the regions containing objects that might have been overlooked by the learned attention representation. Through extensive evaluation on VQAv1, VQAv2 and TDIUC datasets, we show that incorporating complementary QAA allows state-of-the-art VQA models to perform better, and provides significant boost to simplistic VQA models, enabling them to performance on par with highly sophisticated fusion strategies., To appear in the proceedings of International Conference on Pattern Recognition (ICPR) 2020

This paper reviewed the 3rd NTIRE challenge on single-image super-resolution (restoration of rich details in a low-resolution image) with a focus on proposed solutions and results. The challenge had 1 track, which was aimed at the real-world single image super-resolution problem with an unknown scaling factor. Participants were mapping low-resolution images captured by a DSLR camera with a shorter focal length to their high-resolution images captured at a longer focal length. With this challenge, we introduced a novel real-world super-resolution dataset (RealSR). The track had 403 registered participants, and 36 teams competed in the final testing phase. They gauge the state-of-the-art in real-world single image super-resolution.

Deep neural networks are vulnerable to adversarial attacks, which can fool them by adding minuscule perturbations to the input images. The robustness of existing defenses suffers greatly under white-box attack settings, where an adversary has full knowledge about the network and can iterate several times to find strong perturbations. We observe that the main reason for the existence of such perturbations is the close proximity of different class samples in the learned feature space. This allows model decisions to be totally changed by adding an imperceptible perturbation in the inputs. To counter this, we propose to class-wise disentangle the intermediate feature representations of deep networks. Specifically, we force the features for each class to lie inside a convex polytope that is maximally separated from the polytopes of other classes. In this manner, the network is forced to learn distinct and distant decision regions for each class. We observe that this simple constraint on the features greatly enhances the robustness of learned models, even against the strongest white-box attacks, without degrading the classification performance on clean images. We report extensive evaluations in both black-box and white-box attack scenarios and show significant gains in comparison to state-of-the art defenses., Comment: Accepted at ICCV 2019

Deep neural networks (DNNs) have shown vulnerability to adversarial attacks, i.e., carefully perturbed inputs designed to mislead the network at inference time. Recently introduced localized attacks, Localized and Visible Adversarial Noise (LaVAN) and Adversarial patch, pose a new challenge to deep learning security by adding adversarial noise only within a specific region without affecting the salient objects in an image. Driven by the observation that such attacks introduce concentrated high-frequency changes at a particular image location, we have developed an effective method to estimate noise location in gradient domain and transform those high activation regions caused by adversarial noise in image domain while having minimal effect on the salient object that is important for correct classification. Our proposed Local Gradients Smoothing (LGS) scheme achieves this by regularizing gradients in the estimated noisy region before feeding the image to DNN for inference. We have shown the effectiveness of our method in comparison to other defense methods including Digital Watermarking, JPEG compression, Total Variance Minimization (TVM) and Feature squeezing on ImageNet dataset. In addition, we systematically study the robustness of the proposed defense mechanism against Back Pass Differentiable Approximation (BPDA), a state of the art attack recently developed to break defenses that transform an input sample to minimize the adversarial effect. Compared to other defense mechanisms, LGS is by far the most resistant to BPDA in localized adversarial attack setting., Accepted At WACV-2019

Current Zero-Shot Learning (ZSL) approaches are restricted to recognition of a single dominant unseen object category in a test image. We hypothesize that this setting is ill-suited for real-world applications where unseen objects appear only as a part of a complex scene, warranting both ‘recognition’ and ‘localization’ of an unseen category. To address this limitation, we introduce a new ‘Zero-Shot Detection’ (ZSD) problem setting, which aims at simultaneously recognizing and locating object instances belonging to novel categories without any training examples. We also propose a new experimental protocol for ZSD based on the highly challenging ILSVRC dataset, adhering to practical issues, e.g., the rarity of unseen objects. To the best of our knowledge, this is the first end-to-end deep network for ZSD that jointly models the interplay between visual and semantic domain information. To overcome the noise in the automatically derived semantic descriptions, we utilize the concept of meta-classes to design an original loss function that achieves synergy between max-margin class separation and semantic space clustering. Furthermore, we present a baseline approach extended from recognition to ZSD setting. Our extensive experiments show significant performance boost over the baseline on the imperative yet difficult ZSD problem.

Being heavily reliant on animals, it is our ethical obligation to improve their well-being by understanding their needs. Several studies show that animal needs are often expressed through their faces. Though remarkable progress has been made towards the automatic understanding of human faces, this has regrettably not been the case with animal faces. There exists significant room and appropriate need to develop automatic systems capable of interpreting animal faces. Among many transformative impacts, such a technology will foster better and cheaper animal healthcare, and further advance animal psychology understanding. We believe the underlying research progress is mainly obstructed by the lack of an adequately annotated dataset of animal faces, covering a wide spectrum of animal species. To this end, we introduce a large-scale, hierarchical annotated dataset of animal faces, featuring 21.9K faces from 334 diverse species and 21 animal orders across biological taxonomy. These faces are captured `in-the-wild' conditions and are consistently annotated with 9 landmarks on key facial features. The proposed dataset is structured and scalable by design; its development underwent four systematic stages involving rigorous, manual annotation effort of over 6K man-hours. We benchmark it for face alignment using the existing art under novel problem settings. Results showcase its challenging nature, unique attributes and present definite prospects for novel, adaptive, and generalized face-oriented CV algorithms. We further benchmark the dataset for face detection and fine-grained recognition tasks, to demonstrate multi-task applications and room for improvement. Experiments indicate that this dataset will push the algorithmic advancements across many related CV tasks and encourage the development of novel systems for animal facial behaviour monitoring. We will make the dataset publicly available., Comment: 15 pages, 14 figures

Recently proposed Capsule Network is a brain inspired architecture that brings a new paradigm to deep learning by modelling input domain variations through vector based representations. Despite being a seminal contribution, CapsNet does not explicitly model structured relationships between the detected entities and among the capsule features for related inputs. Motivated by the working of cortical network in HVS, we seek to resolve CapsNet limitations by proposing several intuitive modifications to the CapsNet architecture. We introduce, (1) a novel routing weight initialization technique, (2) an improved CapsNet design that exploits semantic relationships between the primary capsule activations using a densely connected Conditional Random Field and (3) a Cholesky transformation based correlation module to learn a general priority scheme. Our proposed design allows CapsNet to scale better to more complex problems, such as the multi-label classification task, where semantically related categories co-exist with various interdependencies. We present theoretical bases for our extensions and demonstrate significant improvements on ADE20K scene dataset.

3D shape generation is a challenging problem due to the high-dimensional output space and complex part configurations of real-world objects. As a result, existing algorithms experience difficulties in accurate generative modeling of 3D shapes. Here, we propose a novel factorized generative model for 3D shape generation that sequentially transitions from coarse to fine scale shape generation. To this end, we introduce an unsupervised primitive discovery algorithm based on a higher-order conditional random field model. Using the primitive parts for shapes as attributes, a parameterized 3D representation is modeled in the first stage. This representation is further refined in the next stage by adding fine scale details to shape. Our results demonstrate improved representation ability of the generative model and better quality samples of newly generated 3D shapes. Further, our primitive generation approach can accurately parse common objects into a simplified representation., Comment: CVPR 2019

In-line with the success of deep learning on traditional recognition problem, several end-to-end deep models for zero-shot recognition have been proposed in the literature. These models are successful to predict a single unseen label given an input image, but does not scale to cases where multiple unseen objects are present. In this paper, we model this problem within the framework of Multiple Instance Learning (MIL). To the best of our knowledge, we propose the first end-to-end trainable deep MIL framework for the multi-label zero-shot tagging problem. Due to its novel design, the proposed framework has several interesting features: (1) Unlike previous deep MIL models, it does not use any off-line procedure (e.g., Selective Search or EdgeBoxes) for bag generation. (2) During test time, it can process any number of unseen labels given their semantic embedding vectors. (3) Using only seen labels per image as weak annotation, it can produce a bounding box for each predicted label. We experiment with large-scale NUS-WIDE dataset and achieve superior performance across conventional, zero-shot and generalized zero-shot tagging tasks.

Background: Bronchial asthma is a social and economic healthcare burden. Drug Utilization studies can play a key role in helping the healthcare system to understand, interpret and improve the prescribing administration & use of medications.Methods: A prospective, Cross-sectional, Observational study was conducted at inpatient and outpatient Department of Medicine and Paediatrics in tertiary health care centre. As per inclusion criteria, 127 patients with asthma were interviewed, and prescription data was recorded in a pre-designed case record form from 1 October 2018 to 31 December 2018.Results: A total of 127 prescriptions were studied. It was observed that majority of the study patients were in the age group of 40-50 years, with 83 (65.5%) males and 44 (34.6%) females. Most of the patients were of Mild intermittent asthma 65 (51.18 %) followed by patients with Mild persistent asthma 31 (24.40 %). 10 different anti-asthmatics drugs were prescribed along with adjunctive medicines. Most of the patients received multiple drug therapy 98 (77.16 %). Prevalence of two drug combination was the highest 74 (58.26 %) followed by monotherapy29 (22.83 %). Most common route of drug administration was inhalational route followed by oral route. β2 agonists and corticosteroids were the most commonly prescribed combination drugs followed by methylxanthines. 103 (81.10 %) prescription contains both generic and brand drugs compared to generic names only 16 (12.59 %) and brand names only 8 (6.29 %).Conclusions: This study concluded that the present prescribing practice in asthma therapy in tertiary care centre is not sufficiently rational. Hence, it is necessary to encourage physicians to follow asthma guidelines while managing asthmatic patients.

Background: The illness due to psychiatric diseases and behavioral disorders is enormous, but it remains under represented by conventional public health statistics; which mainly targets mortality more than morbidity or dysfunction. The drug utilization study was conducted to find out problems and provide feedback to prescribers so as to create awareness about irrational use of drugs. This study was conducted in psychiatry outpatient department (OPD) of a tertiary care hospital.Methods: A prospective, observational and cross-sectional study was conducted in Department of Pharmacology in collaboration with department of psychiatry in a tertiary care hospital. Patients fulfilling the inclusion criteria visiting psychiatric OPD from 01 October 2018 to 31 December 2018 period of 3 months were enrolled in the study. Assessment of the prescriptions was done as per the ICD-10 criteria and World Health Organization-indicators for rational use of drugs (WHO-INRUD).Results: Among total 390 prescriptions; 71.54% was males and 28.46% were females. All the drugs were oral formulations. The burden of diseases are depression 35.89%, anxiety 20.51%, schizophrenia 18.97% and bipolar disorder 11.28% and other psychiatric diseases were 13.33%. Olanzapine was the most common drug which was prescribed followed by alprazolam. Total 1,140 drugs were prescribed out of which 86.32% was psychotropic drug and 13.68% were other like vitamin B-complex, analgesics, anta-acids, antibiotics and NSAIDS. The Most common prescribed class was atypical antipsychotics 26.31% followed by benzodiazepines 20.26%, antipsychotics 19.74%, tricyclic antidepressants 8.16%, anticonvulsants 9.21%, antiepileptics 2.37% and lastly selective serotonin reuptake inhibitors.Conclusions: Atypical antipsychotics was most commonly prescribed drug followed by benzodiazepines. All the drugs were prescribed by their generic names which suggests awareness among prescribers regarding WHO policies for generic prescribing.

Background: Drug utilization study is of utmost importance in the elderly. The consumption of drug amongst elderly segment of society is maximum and many of them use at least three prescribed drugs, concurrently. One of the plausible explanation of usage of large number of medicines is prevalence of multi-morbidities as well as suffer from chronic and degenerative pathology amongst them along with the alteration of pharmacokinetics and pharmacodynamics of many drugs are altered with advanced age. Thus, the aim of this study was to assess current prescribing practices among physicians for geriatric patients in tertiary care hospital.Methods: A prospective, observational and cross-sectional study was conducted in Department of Pharmacology in collaboration with Department of Medicine in a tertiary healthcare hospital. As per inclusion criteria, patients visiting Medicine outpatient department (OPD) from 01 October 2018 to 31 December 2018 were enrolled in the study.Results: The average number of drugs per encounter was 4.1. 100% of the drugs were prescribed by their generic name and those prescribed from World Health Organization (WHO) list of essential medicines were 70.2%. Analysis of polypharmacy showed maximum 11 and minimum 1 drug was prescribed. Overall analysis of the drugs showed, the most common route of administration was orally. Drugs acting on cardiovascular system (CVS), gastrointestinal system (GI) system, analgesics and anti-inflammatory drugs and vitamins and mineral supplements constitute the major bulk of drug prescribed.Conclusions: In our study, though polypharmacy was found but it was necessary unavoidable, but was within the WHO standards of 1.6 to 4.8. Drugs acting on CVS and GI system were the most common drugs prescribed. Prescriptions with injections and steroids were less in the study population. All the drugs were prescribed by their generic name and most were present in the essential list of medicines.

We present a framework to automatically detect and remove shadows in real world scenes from a single image. Previous works on shadow detection put a lot of effort in designing shadow variant and invariant hand-crafted features. In contrast, our framework automatically learns the most relevant features in a supervised manner using multiple convolutional deep neural networks (ConvNets). The features are learned at the super-pixel level and along the dominant boundaries in the image. The predicted posteriors based on the learned features are fed to a conditional random field model to generate smooth shadow masks. Using the detected shadow masks, we propose a Bayesian formulation to accurately extract shadow matte and subsequently remove shadows. The Bayesian formulation is based on a novel model which accurately models the shadow generation process in the umbra and penumbra regions. The model parameters are efficiently estimated using an iterative optimization procedure. Our proposed framework consistently performed better than the state-of-the-art on all major shadow databases collected under a variety of conditions.

Current Visual Question Answering (VQA) systems can answer intelligent questions about ‘known’ visual content. However, their performance drops significantly when questions about visually and linguistically ‘unknown’ concepts are presented during inference (‘Open-world’ scenario). A practical VQA system should be able to deal with novel concepts in real world settings. To address this problem, we propose an exemplar-based approach that transfers learning (i.e., knowledge) from previously ‘known’ concepts to answer questions about the ‘unknown’. We learn a highly discriminative joint embedding (JE) space, where visual and semantic features are fused to give a unified representation. Once novel concepts are presented to the model, it looks for the closest match from an exemplar set in the JE space. This auxiliary information is used alongside the given Image-Question pair to refine visual attention in a hierarchical fashion. Our novel attention model is based on a dual-attention mechanism that combines the complementary effect of spatial and channel attention. Since handling the high dimensional exemplars on large datasets can be a significant challenge, we introduce an efficient matching scheme that uses a compact feature description for search and retrieval. To evaluate our model, we propose a new dataset for VQA, separating unknown visual and semantic concepts from the training set. Our approach shows significant improvements over state-of-the-art VQA models on the proposed Open-World VQA dataset and other standard VQA datasets.

Background: With the surge of basic and clinical research activities, and subsequent recognition of newer dermatological disorders; dermatology is now considered to be one of the most important specialties in a healthcare setup. Dermatological problem in India manifests as primary and secondary cutaneous complaints. The aim of this study was to assess current prescribing practices among physicians for dermatology patients in tertiary care hospital.Methods: A prospective, observational and cross-sectional study was conducted in Department of Pharmacology in collaboration with Department of Dermatology in a tertiary healthcare hospital. As per inclusion criteria, patients visiting Dermatology Outpatient department (OPD) from October 2018 to December 2018 were enrolled in the study.Results: The average number of drugs per encounter was 4.9. 100% of the drugs were prescribed by their generic name and those prescribed from World Health Organization (WHO) list of essential medicines were 68.9%. Analysis of polypharmacy showed maximum 11 and minimum 2 drug was prescribed. Overall analysis of the drugs showed, the most common route of administration was orally. Antihistamines, antifungals and steroids constitute the major bulk of drug prescribed.Conclusions: In our study, though polypharmacy was found but it was necessary as per the conditions. Anti-fungal, antihistamines and steroids were the most common drugs prescribed. Prescriptions with injections and steroids were less in the study population. All the drugs were prescribed by their generic name and most were present in the essential list of medicines. Use of Nonsteroidal anti-inflammatory drugs (NSAIDs) was higher in our study as compared to other studies.

Background: Hypertension is the most common modifiable risk factor for cardiovascular diseases, stroke and renal failure. Drug utilization research facilitates the rational use of drugs and provides an insight into the pattern of drug use. As such, the present study was conducted to assess the prescribing patterns of antihypertensive agents and to observe the disparity between the recommended and actual practices for pharmacological management of hypertension at a tertiary care hospital.Methods: A prospective, observational and cross-sectional study was conducted in Department of Pharmacology in collaboration with Department of Medicine in a tertiary healthcare hospital. As per inclusion criteria, patients with hypertension visiting Medicine OPD from 1st October 2018 to 31st December 2018 were enrolled in the study.Results: A total of 390 prescriptions were collected. It was observed that majority of the study patients were female (53.84%) and were >60 years of age. Average number of drugs prescribed per prescription was 4.18. Multiple drug regimen (67.69%) was the mainstay therapy in our study. Prevalence of two drug combination was the highest (46.92%) followed by monotherapy (32.3%). Calcium channel blockers (CCBs) (amlodipine) was the most frequently prescribed class in monotherapy (19.23%). In two drug regimens, the combination of CCB with angiotensin-converting enzyme inhibitors (ACEIs) (combination of amlodipine and enalapril) was prescribed the most (28.46%) whereas prescription of combination of ACEI, beta-blocker (BB) and CCB (enalapril, atenolol and amlodipine combination) was the highest (13.84%) in three drug regimens.Conclusions: The prescription pattern of antihypertensive drugs was found to be rational and mainly in accordance with Joint National Committee VIII guidelines. All the drugs were prescribed by their generic names which suggests awareness among prescribers regarding WHO policies for generic prescribing.

The reduction of electric power consumption is considered to be a critical issue especially in the areas with elevated temperatures such as Iraq. In Basra city, the electric power consumption due to the air conditioning units (ACU) is so high. In this work, an evaporative cooling unit (ECU) is fabricated and tested where it is retrofitted with the condenser part of a split-type air conditioner. The experimental procedure was done in Basra engineering technical college located in west of Basra city. The present work is categorized in two cases. In case one, the system runs without using the ECU, while in case two the system is tested with ECU. The obtained results show that decreasing the inlet air temperature to the condenser will drop off the condensation and evaporation temperatures. Besides, it decreases the compression pressure ratio and increases the refrigerant mass flow rate. With this modification, it is found that the consumption power is decreased by 26% and COP increased by 90% compared with case one. The second law analysis of the modified system shows that using ECU will decrease most the exergy destruction and losses in the components. Only exergy destruction in the evaporator is increased. The system second law efficiency is improved by 69.5%.

Variational auto-encoders (VAEs) provide an attractive solution to image generation problem. However, they tend to produce blurred and over-smoothed images due to their dependence on pixel-wise reconstruction loss. This paper introduces a new approach to alleviate this problem in the VAE based generative models. Our model simultaneously learns to match the data, reconstruction loss and the latent distributions of real and fake images to improve the quality of generated samples. To compute the loss distributions, we introduce an auto-encoder based discriminator model which allows an adversarial learning procedure. The discriminator in our model also provides perceptual guidance to the VAE by matching the learned similarity metric of the real and fake samples in the latent space. To stabilize the overall training process, our model uses an error feedback approach to maintain the equilibrium between competing networks in the model. Our experiments show that the generated samples from our proposed model exhibit a diverse set of attributes and facial expressions and scale up to highresolution images very well.

Inferring the correspondences between consecutive video frames with high accuracy is essential for many medical image processing and computer vision tasks (e.g. image mosaicking, 3D scene reconstruction). Image correspondences can be computed by feature extraction and matching algorithms, which are computationally expensive and are challenged by low texture frames. Convolutional neural networks (CNN) can estimate dense image correspondences with high accuracy, but lack of labeled data especially in medical imaging does not allow end-to-end supervised training. In this paper, we present an unsupervised learning method to estimate dense image correspondences (DIC) between endoscopy frames by developing a new CNN model, called the EndoRegNet. Our proposed network has three distinguishing aspects: a local DIC estimator, a polynomial image transformer which regularizes local correspondences and a visibility mask which refines image correspondences. The EndoRegNet was trained on a mix of simulated and real endoscopy video frames, while its performance was evaluated on real endoscopy frames. We compared the results of EndoRegNet with traditional feature-based image registration. Our results show that EndoRegNet can provide faster and more accurate image correspondences estimation. It can also effectively deal with deformations and occlusions which are common in endoscopy video frames without requiring any labeled data.

Inexpensive structured light sensors can capture rich information from indoor scenes, and scene labeling problems provide a compelling opportunity to make use of this information. In this paper we present a novel conditional random field (CRF) model to effectively utilize depth information for semantic labeling of indoor scenes. At the core of the model, we propose a novel and efficient plane detection algorithm which is robust to erroneous depth maps. Our CRF formulation defines local, pairwise and higher order interactions between image pixels. At the local level, we propose a novel scheme to combine energies derived from appearance, depth and geometry-based cues. The proposed local energy also encodes the location of each object class by considering the approximate geometry of a scene. For the pairwise interactions, we learn a boundary measure which defines the spatial discontinuity of object classes across an image. To model higher-order interactions, the proposed energy treats smooth surfaces as cliques and encourages all the pixels on a surface to take the same label. We show that the proposed higher-order energies can be decomposed into pairwise sub-modular energies and efficient inference can be made using the graph-cuts algorithm. We follow a systematic approach which uses structured learning to fine-tune the model parameters. We rigorously test our approach on SUN3D and both versions of the NYU-Depth database. Experimental results show that our work achieves superior performance to state-of-the-art scene labeling techniques.

Spectral signatures of natural scenes were earlier found to be distinctive for different scene types with varying spatial envelope properties such as openness, naturalness, ruggedness, and symmetry. Recently, such handcrafted features have been outclassed by deep learning based representations. This paper proposes a novel spectral description of convolution features, implemented efficiently as a unitary transformation within deep network architectures. To the best of our knowledge, this is the first attempt to use deep learning based spectral features explicitly for image classification task. We show that the spectral transformation decorrelates convolutional activations, which reduces co-adaptation between feature detections, thus acts as an effective regularizer. Our approach achieves significant improvements on three large-scale scene-centric datasets (MIT-67, SUN-397, and Places-205). Furthermore, we evaluated the proposed approach on the attribute detection task where its superior performance manifests its relevance to semantically meaningful characteristics of natural scenes.

Conventional change detection methods require a large number of images to learn background models or depend on tedious pixel-level labeling by humans. In this paper, we present a weakly supervised approach that needs only image-level labels to simultaneously detect and localize changes in a pair of images. To this end, we employ a deep neural network with DAG topology to learn patterns of change from image-level labeled training data. On top of the initial CNN activations, we define a CRF model to incorporate the local differences and context with the dense connections between individual pixels. We apply a constrained mean-field algorithm to estimate the pixel-level labels, and use the estimated labels to update the parameters of the CNN in an iterative EM framework. This enables imposing global constraints on the observed foreground probability mass function. Our evaluations on four benchmark datasets demonstrate superior detection and localization performance.

Innovative advance dielectric materials were utilized to fabricate novel thin film capacitor device. This includes ceramic and polymeric material combination which is considered to enhance the dielectric polarization. The dielectric film presented in the study focused on PbTiO3/PVDF-TrFE which was prepared through layer-by-layer spin coating deposition method. The PVDF-TrFE coating was aimed to encapsulate the fragility and to improve the crystallinity property of PbTiO3 thin film surface. The preferential orientation of (1 0 1) tetragonal perovskite PbTiO3 was obtained in this and high crystallinity was found on bilayered PbTiO3/PVDF-TrFE film. The bilayered film produced good film surface in term of its uniformity and surface roughness. The er and tan (δ) value of bilayered PbTiO 3 / PVDF-TrFE film was taken at 1 kHz that approximately 220 and 0.09 respectively.

Recent advances in deep learning have resulted in human-level performances on popular unconstrained face datasets including Labeled Faces in the Wild and YouTube Faces. To further advance research, IJB-A benchmark was recently introduced with more challenges especially in the form of extreme head poses. Registration of such faces is quite demanding and often requires laborious procedures like facial landmark localization. In this paper, we propose a Convolutional Neural Networks based data-driven approach which learns to simultaneously register and represent faces. We validate the proposed scheme on template based unconstrained face identification. Here, a template contains multiple media in the form of images and video frames. Unlike existing methods which synthesize all template media information at feature level, we propose to keep the template media intact. Instead, we represent gallery templates by their trained one-vs-rest discriminative models and then employ a Bayesian strategy which optimally fuses decisions of all medias in a query template. We demonstrate the efficacy of the proposed scheme on IJB-A, YouTube Celebrities and COX datasets where our approach achieves significant relative performance boosts of 3.6%, 21.6% and 12.8% respectively.