Water causes degradation of quality in optical images captured underwater due to its physical properties of absorption and scattering. This degradation is further aggravated by the increase in water depth and the presence of contaminated water. Transformers in the vision domain have made a quantum leap in many vision tasks such as detection, and segmentation but yet to make any progress in enhancing degraded underwater images. We propose a transformer-based model named “Aquaformer” which makes four major contributions: an adaptive layer normalization, replacement of masked cyclic shift with symmetric padding in window partitioning, a novel aggregation mechanism, and an adjustable fusion approach. These succeed in making the model a very powerful one, producing significantly better performance compared to the latest state-of-the-art methods. Testing on multiple benchmark datasets, employing both quantitative and qualitative metrics, establishes its supremacy.

In the past few decades, due to rapid growth in industrialization, there has been a steady decline of the air quality along with an increase in the concentration of PM2.5. It is well known that a high PM2.5 concentration adversely affects the environment and has hazardous impact on public health. Therefore, it is important to monitor the PM2.5 concentration at geographic locations where air quality monitoring stations are presently unavailable, especially in remote areas. Unfortunately, installation of such monitoring stations requires expensive instruments and constant maintenance. This paper presents a novel, low-cost and portable alternative to such measurement apparatus, where PM2.5 concentration is estimated based on image input obtained from a camera. The novelty of the present work lies in its hitherto unique attempt to capture information regarding PM2.5 content from visibility degradation caused by the pollutant which is further supplemented by important knowledge regarding seasonal and diurnal variation of it. The latter has a crucial role in the prevention of confounding effects arising from the presence of other weather and atmospheric elements. Another important highlight is the use of a full reference image metric as a feature, for which a powerful, dehazing algorithm has been employed. The results obtained are extremely promising, providing a close to accurate estimation of PM2.5 concentration with R2values far higher than reported in the literature. To summarize, the construction of a unique feature set, together with an appropriate machine learning algorithm, lead to an extremely reliable, stand-alone approach, deployable on a hand-held device such as a mobile and is a very significant contribution indeed of the proposed approach.

With the emergence of remote storage, computation facilities, and the availability of high-speed data connectivity — cloud computation has become the call for the day. In this scenario, security and computability of data have emerged as two crucial aspects which often conflict with each other. An efficient solution is to build a trade-off between the two. In this work, we propose a novel Hilbert-Curve-based data encoding scheme (HC encoding) which obfuscates the data points. We use the obfuscated data in DBSCAN clustering and obtain an approximated output with respect to the original clustering (without invoking the plaintext data at any stages of the computation) . HC encoding has a provision to modulate the degree of obfuscation. The correctness of the clustering performance decreases with the increase in obfuscation and vice versa. The empirical study is carried out on 8 regular datasets and a large, color image dataset, CIFAR10. Two key applications are explored– (i) clustering of homogeneous images and (ii) clustering of heterogeneous images originating from two different sources. The empirical results show the effectiveness of the HC encoding to provide approximate clustering performances on the obfuscated multimedia data as well as regular datasets. The outcomes also manifest HC encoding’s superiority against the comparing methods which operate under a similar fully private constraint.

There is often a mixture of very frequent labels and very infrequent labels in multi-label datatsets. This variation in label frequency, a type class imbalance, creates a significant challenge for building efficient multi-label classification algorithms. In this paper, we tackle this problem by proposing a minority class oversampling scheme, UCLSO, which integrates Unsupervised Clustering and Label-Specific data Oversampling. Clustering is performed to find out the key distinct and locally connected regions of a multi-label dataset (irrespective of the label information). Next, for each label, we explore the distributions of minority points in the cluster sets. Only the minority points within a cluster are used to generate the synthetic minority points that are used for oversampling. Even though the cluster set is the same across all labels, the distributions of the synthetic minority points will vary across the labels. The training dataset is augmented with the set of label-specific synthetic minority points, and classifiers are trained to predict the relevance of each label independently. Experiments using 12 multi-label datasets and several multi-label algorithms show that the proposed method performed very well compared to the other competing algorithms.

Rapid snow melting and consequent decrease in the snow cover of a particular location plays a significant role in the occurrence of glacial lake outburst floods (GLOFs). Therefore, determination of the rate of depletion in the snow covered area (SCA) and understanding the underlying causes is of utmost importance for early detection of such disasters. This work employs cellular automata (CA) for predicting the spatio-temporal change of SCA in and around the Shishper glacier, Baltistan, which has experienced recurrent annual occurrences of GLOFs. The CA forecasts the future state of a cell, taking into account it’s current snow cover as well as terrain and climatic factors. The major contribution of this work is in the formulation of a unique ternary CA model, with the 3-classes being snow, land, and lake, which effortlessly captures the decrease in SCA and concomitant increase in lake area with a notable prediction accuracy of 84%.

Abstract In this article, we present a novel neighbourhood based multi‐label classifier, Multi‐label Learning on principles of Reverse k‐Nearest Neighbourhood (ML‐RkNN) where we estimate the neighbourhood of the points on the basis of their reverse k‐nearest neighbourhood (RkNN). Through RkNN, for the same value of k, we get different number of neighbours for different instances and this happens adaptively according to the neighbourhood configuration of the points. The automatically adaptive neighbourhood helps us in better learning of the local configurations around the points. Our scheme also facilitates implicit handling of the local imbalances prevailing in the datasets by comparing the class distributions of the test points and their reverse nearest neighbours. This implicit and adaptive handling is particularly useful for multi‐label label datasets, whose labels are differentially imbalanced. Empirical study is performed on 10 real‐world multi‐label datasets considering five neighbourhood based multi‐label learners. Macro F1 is used as the evaluating metric. The proposed method has given statistically superior and statistically comparable performances with respect to three and two comparing methods respectively. Additionally, we have explored the use of two different distance metrics, Euclidean and Jaccard in our scheme for nominal datasets.

This paper presents a novel approach to the extraction of weaker periodic signals in the presence of noise and a stronger periodic waveform. The signals may have any shape, not necessarily sinusoidal in nature. The approach is based on the Singular Value Decomposition. The extraction procedure consists of two steps, identifying the strongest periodic component and consequent configuring the data matrix followed by extraction of this component. This process is repeated till all the components have been recovered. The approach is both simple and robust and requires no additional reference inputs. The simplicity of the scheme is however, at the expense of certain inherent limitations, which are also investigated.>

This paper considers the problem of estimating the direction of arrival of plane waves from sensor array data. It is shown through a multivariate generalization of Whittle's inequality that the least squares estimator and its subspace-based approximations are inefficient when the noise distribution is non-Gaussian. In order to improve efficiency, two computationally elegant extensions of the MUSIC algorithm are suggested. Computer simulations indicate that the proposed methods perform much better than MUSIC for non-Gaussian noise, while the performances of all the three methods are comparable for Gaussian noise.>