Understanding and predicting the human visual attention mechanism is an active area of research in the fields of neuroscience and computer vision. In this paper, we propose DeepFix, a fully ...convolutional neural network, which models the bottom-up mechanism of visual attention via saliency prediction. Unlike classical works, which characterize the saliency map using various hand-crafted features, our model automatically learns features in a hierarchical fashion and predicts the saliency map in an end-to-end manner. DeepFix is designed to capture semantics at multiple scales while taking global context into account, by using network layers with very large receptive fields. Generally, fully convolutional nets are spatially invariant-this prevents them from modeling location-dependent patterns (e.g., centre-bias). Our network handles this by incorporating a novel location-biased convolutional layer. We evaluate our model on multiple challenging saliency data sets and show that it achieves the state-of-the-art results.
Deep convolutional neural networks (CNNs) have revolutionized the computer vision research and have seen unprecedented adoption for multiple tasks, such as classification, detection, and caption ...generation. However, they offer little transparency into their inner workings and are often treated as black boxes that deliver excellent performance. In this paper, we aim at alleviating this opaqueness of CNNs by providing visual explanations for the network's predictions. Our approach can analyze a variety of CNN-based models trained for computer vision applications, such as object recognition and caption generation. Unlike the existing methods, we achieve this via unraveling the forward pass operation. The proposed method exploits feature dependencies across the layer hierarchy and uncovers the discriminative image locations that guide the network's predictions. We name these locations CNN fixations, loosely analogous to human eye fixations. Our approach is a generic method that requires no architectural changes, additional training, or gradient computation, and computes the important image locations (CNN fixations). We demonstrate through a variety of applications that our approach is able to localize the discriminative image locations across different network architectures, diverse vision tasks, and data modalities.
Machine learning models are susceptible to adversarial perturbations: small changes to input that can cause large changes in output. It is also demonstrated that there exist input-agnostic ...perturbations, called universal adversarial perturbations, which can change the inference of target model on most of the data samples. However, existing methods to craft universal perturbations are (i) task specific, (ii) require samples from the training data distribution, and (iii) perform complex optimizations. Additionally, because of the data dependence, fooling ability of the crafted perturbations is proportional to the available training data. In this paper, we present a novel, generalizable and data-free approach for crafting universal adversarial perturbations. Independent of the underlying task, our objective achieves fooling via corrupting the extracted features at multiple layers. Therefore, the proposed objective is generalizable to craft image-agnostic perturbations across multiple vision tasks such as object recognition, semantic segmentation, and depth estimation. In the practical setting of black-box attack scenario (when the attacker does not have access to the target model and it's training data), we show that our objective outperforms the data dependent objectives to fool the learned models. Further, via exploiting simple priors related to the data distribution, our objective remarkably boosts the fooling ability of the crafted perturbations. Significant fooling rates achieved by our objective emphasize that the current deep learning models are now at an increased risk, since our objective generalizes across multiple tasks without the requirement of training data for crafting the perturbations. To encourage reproducible research, we have released the codes for our proposed algorithm.1
We introduce a detection framework for dense crowd counting and eliminate the need for the prevalent density regression paradigm. Typical counting models predict crowd density for an image as opposed ...to detecting every person. These regression methods, in general, fail to localize persons accurate enough for most applications other than counting. Hence, we adopt an architecture that locate s every person in the crowd, size s the spotted heads with bounding box and then count s them. Compared to normal object or face detectors, there exist certain unique challenges in designing such a detection system. Some of them are direct consequences of the huge diversity in dense crowds along with the need to predict boxes contiguously. We solve these issues and develop our LSC-CNN model, which can reliably detect heads of people across sparse to dense crowds. LSC-CNN employs a multi-column architecture with top-down feature modulation to better resolve persons and produce refined predictions at multiple resolutions. Interestingly, the proposed training regime requires only point head annotation, but can estimate approximate size information of heads. We show that LSC-CNN not only has superior localization than existing density regressors, but outperforms in counting as well. The code for our approach is available at https://github.com/val-iisc/lsc-cnn .
We propose a novel crowd counting model that maps a given crowd scene to its density. Crowd analysis is compounded by myriad of factors like inter-occlusion between people due to extreme crowding, ...high similarity of appearance between people and background elements, and large variability of camera view-points. Current state-of-the art approaches tackle these factors by using multi-scale CNN architectures, recurrent networks and late fusion of features from multi-column CNN with different receptive fields. We propose switching convolutional neural network that leverages variation of crowd density within an image to improve the accuracy and localization of the predicted crowd count. Patches from a grid within a crowd scene are relayed to independent CNN regressors based on crowd count prediction quality of the CNN established during training. The independent CNN regressors are designed to have different receptive fields and a switch classifier is trained to relay the crowd scene patch to the best CNN regressor. We perform extensive experiments on all major crowd counting datasets and evidence better performance compared to current state-of-the-art methods. We provide interpretable representations of the multichotomy of space of crowd scene patches inferred from the switch. It is observed that the switch relays an image patch to a particular CNN column based on density of crowd.
A race to achieve a crossover from positive to negative magnetoresistance is intense in the field of nanostructured materials to reduce the size of memory devices. Here, the unusual complex ...magnetoresistance in nonmagnetic sulfur‐doped Sb2Se3 nanowires is demonstrated. Intentionally, sulfur is doped in such a way to nearly achieve the charge neutrality point that is evident from switching of carrier type from p‐type to n‐type at 13 K as inferred from the low‐temperature thermoelectric power measurements. A change from 3D variable range hopping (VRH) to power law transport with α = 0.18 in resistivity measurement signifies a Luttinger liquid transport with weak links through the nanowires. Interestingly, high magnetic field induced negative magnetoresistance (NMR) occurring in hole dominated temperature regimes can only be explained by invoking the concept of charge puddles. Spot energy dispersive spectroscopy (EDS), magnetic force microscopy (MFM) measurements, Tmott and Regel plot indicate an enhanced disorder in these sulfurized nanowires that are found to be the precursor for the formation of these charge puddles. Tunability of conducting states in these nanowires is investigated in the light of interplay of carrier type, magnetic field, temperature, and intricate intra‐inter wire transport that makes this nanowires potential for large scale spintronic devices.
A complex crossover in magnetoresistance is observed and investigated in sulfur‐doped nanowires synthesized by microwave assisted solvothermal reaction. The role of charge puddles and tunability of conducting states in these nanowires is discussed as an interplay between carrier type, magnetic field, temperature, and intricate intra‐inter wire transport that makes these nanowires potentially suitable for large‐scale spintronic devices.
In this review, we examine the roles of G protein-coupled receptors (GPCRs) as noncanonical pattern recognition receptors (PRRs) and also discuss how GPCR signaling in C. elegans regulates various ...immune processes. The search for PRRs and immune regulators in C. elegans prompted many research groups on a quest to identify neuronal GPCRs or GPCR signaling components that are involved in immune regulation. AMP, antimicrobial peptides; DAMP, damage-associated molecular pattern; GPCR, G protein-coupled receptor; HPLA, 4-hydroxyphenyllactic acid; UPR, unfolded protein response. https://doi.org/10.1371/journal.ppat.1009151.g001 A recent study illustrated the role for yet another neuronal GPCR in the regulation of innate immunity by modulating intestinal p38/MAPK activity. ...far, FSHR-1 remains the only intestinal GPCR with an immunomodulatory function.
When it comes to modern death rates, heart disease ranks high. Clinical data analysis has a significant difficulty in the domain of heart disease prediction. For the healthcare business, which ...generates vast amounts of data, Machine Learning (ML) has proven to be a useful tool for aiding in decision-making and prediction. Recent innovations in various domains of the Internet of Things have also made use of ML approaches. Predicting heart disease using ML approaches is only partially explored in the available research. Therefore, to accurately predict heart illness, we provide a unique Multi-Gradient Boosted Adaptive Support Vector Machine (MBASVM). The Multi gradient boosted adaptive support vector machine (MBASVM), an ensemble meta-algorithm, successfully transforms weak learners into strong learners while removing dataset biases for machine learning algorithms. The boosting approach tries to improve the predictability of cardiac disease. For extracting useful features from data, Kernel-Based Principal Component Analysis (K-PCA) is used. The suggested model’s retrieved data are narrowed down using the Chi-Squared Ranker Search (CRS) approach. Measures of recall, sensitivity, specificity, f1-score, accuracy, and precision are used to evaluate the effectiveness of the suggested technique.Comprehensive testing shows that, when compared to other ways, our methodology performed the best.
A class of recent approaches for generating images, called Generative Adversarial Networks (GAN), have been used to generate impressively realistic images of objects, bedrooms, handwritten digits and ...a variety of other image modalities. However, typical GAN-based approaches require large amounts of training data to capture the diversity across the image modality. In this paper, we propose DeLiGAN - a novel GAN-based architecture for diverse and limited training data scenarios. In our approach, we reparameterize the latent generative space as a mixture model and learn the mixture models parameters along with those of GAN. This seemingly simple modification to the GAN framework is surprisingly effective and results in models which enable diversity in generated samples although trained with limited data. In our work, we show that DeLiGAN can generate images of handwritten digits, objects and hand-drawn sketches, all using limited amounts of data. To quantitatively characterize intra-class diversity of generated samples, we also introduce a modified version of inception-score, a measure which has been found to correlate well with human assessment of generated samples.
•Sulfur doped Sb2Se3 nanotubes synthesized by microwave assisted solvothermal method.•Complex crossover of positive to negative magnetoresistance in low magnetic field.•Magnetoresistance analyzed by ...invoking Bipolaron & Wave function shrinkage models.•Thermoelctric: P-type to n-type carrier switching with minority carrier injection.•Potential application in large scale power electronics, spintronics & memory devices.
Materials showing low field induced switch over in magnetoresistance is investigated heavily nowadays for promising power electronics and phase change memory applications. In this work, morphological, topographical, magnetoresistance (MR) and thermoelectric properties of nanobulk assembly of sulfur doped Sb2Se3nanotubes synthesized by novel microwave assisted solvothermal process is investigated and reported. Structural properties indicate that the sample crystallizes in orthorhombic phase while morphology study confirms the formation of nanotubes of length 1–5 μm and diameter 100–150 nm. Electrical resistivity measurements indicates 2D variable range hopping at low temperature (2–50 K), nearest neighbor hopping at intermediate temperature (50–200 K) and thermal activation behavior with activation energy of 2.3 meV from 200 K to 300 K. Interestingly the positive magnetoresistance observed in low magnetic field and at constant low temperatures makes a transition to negative magnetoresistance above 20 K. Bipolaron model along with wave function shrinkage models were invoked to understand the MR measurements while low temperature thermoelectric measurement univocally elucidates that the carrier type switching at 13 K dictates the conduction mechanism as demonstrated by the complex crossover in magnetoresistance.