Showing papers by "International Institute of Information Technology, Hyderabad published in 2023"

A Dynamic Time-Bound Access Control for Secure Hierarchical Content Sharing

Abstract: <p>With rapid growth of mobile users, protecting content from unauthorized users become a complex problem. The concept of temporal role-based access control reduces complexity of user management and restricts access to specified time-slots. But, content privacy is still questionable in case of system resources compromise unexpectedly. Therefore, cryptographic solution for time-bound hierarchical content management is an emerging problem. Most of the related schemes focused on individual user keys and/or revocation, but not on time-bound keys. Hence, these are not well suitable for subscription-based services like pay-TV and newspaper. In this paper, we propose a cryptographic time-bound access control with constant size time-bound keys. In our scheme, subscribed time-slots embed into individual user keys to avoid periodical broadcasting of temporal keys. We prove that our scheme is selectively secure against collision attack under chosen-ciphertext attack. We then discuss cloud-based application to show the strategies of efficient revocation and reduced computational overheads for end-users. </p>

Influence of Vertical Seismic Coefficient in Seismic Analysis of Hydro-Tunnel in Rock

Abstract: As a result of increased urbanization and the need for better infrastructure, the scope of tunnel projects is also expanding. The country has much potential for hydropower, and the hydropower sector has produced the most tunnels. Hydro projects have some of the country's longest tunnels. In the Himalayan areas, almost 75% of the full potential for hydropower production is concentrated. Himachal Pradesh has the most projects with a total tunnel length of 500 km, followed by Uttarakhand (160.8 km) and Jammu and Kashmir (135.14). Due to geographical and geological difficulties, tunneling in the Himalayan area is extremely difficult. Due to geological issues such as sheared rock, high water intrusion, and high geothermal gradient, long hydro-tunnels have experienced time and expense overruns. Extensive studies must be conducted to use the hydropower potential fully, the appropriate technique must be adopted, and risks must be adequately identified and managed. Geological prediction during tunneling should be standard practice to reduce geological uncertainty and prevent unanticipated hazards. To complete a project on schedule and safely accurate evolution, analysis, and interpretation of rock mass quality play a key role. This paper provides a numerical analysis of the seismic response of a circular lined tunnel running through a jointed rock mass. The effect of tunnel depth, frequency, and peak ground acceleration on the axial force produced in the tunnel liner is studied. The outcomes of the numerical computation have verified the patterns of seismic damage observed in the past.

Balanced Sampling-Based Active Learning for Object Detection

Abstract: Active learning focuses on building competitive models using less data by utilizing intelligent sampling, thereby reducing the effort and cost associated with manual data annotation. In this paper an active learning method based on a balanced sampling of images with high and low confidence object detection scores for training an object detector is proposed. Images with higher object prediction scores are sampled using the uncertainty measure proposed by Yu et al. which utilizes detection, classification and distribution statistics. Though this method encourages balanced distribution in sampling, a deeper look into the sampled distribution reveals that the under-represented classes in the initial labeled pool remain skewed throughout the subsequent active learning cycles. To mitigate this problem, in each active learning cycle, we propose to sample an equal proportion of images with high and low confidence object prediction scores from the model trained in the last cycle, where the low confidence prediction sample selection is based on the model’s prediction scores. Experiments conducted on the UEC Food 100 dataset show that the proposed method performs better than the baseline random sampling, CALD and low confidence prediction sampling method by +4.7, +8.7, and +3.1 mean average precision (mAP), respectively. Moreover, consistently superior performance of the proposed method is also demonstrated on the PASCAL VOC’07 and PASCAL VOC’12 datasets.

A Method for Price Prediction of Potato Using Deep Learning Techniques

Abstract: The Indian agricultural sector contributes to 19.9% of the total economy. Agricultural growth can end extreme poverty in a country like India where 70% of the rural population depends on agriculture for their livelihood. It is important to make timely decisions to grow and sustain profitably. Among agricultural items, vegetables have the most supply and price variations. Vegetable prices play a major role in the national economy. It’s tough to keep the supply and prices of vegetables stable because they’re cultivated outside and their yields fluctuate a lot depending on the weather. Despite the government’s efforts to stabilize vegetable supply and pricing, frequent meteorological shifts in recent years have resulted in unpredictable vegetable supply and price swings. Potatoes are one of India’s most popular vegetables. Variation in potato output over time leads to wide price fluctuations, putting growers in a high-risk situation. Therefore, accurate forecasting of prices is critical. The current study tried to investigate variations in potato arrivals and prices in Uttar Pradesh’s biggest potato-producing district, Agra. The data set consists of potato prices from the Achenra market of Agra for the years 2018–2022. The prediction was made and the RMSE value is compared between three different methods namely: Auto-Regressive Integrated Moving Average Model (ARIMA), Artificial Neural Network (ANN) and Long-Short Term Memory (LSTM). When these Rmse values were compared ANN had a lower Rmse value depicting that ANN was best suited for this dataset.

Molecular Dynamics and Machine Learning Study of Adrenaline Dynamics in the Binding Pocket of GPCR

Abstract: G-protein coupled receptors (GPCRs) are the most prominent family of membrane proteins that serve as major targets for one-third of the drugs produced. A detailed understanding of the molecular mechanism of drug-induced activation and inhibition of GPCRs is crucial for the rational design of novel therapeutics. The binding of the neurotransmitter adrenaline to the β2-adrenergic receptor (β2AR) is known to induce a flight or fight cellular response, but much remains to be understood about binding-induced dynamical changes in β2AR and adrenaline. In this article, we examine the potential of mean force (PMF) for the unbinding of adrenaline from the orthosteric binding site of β2AR and the associated dynamics using umbrella sampling and molecular dynamics (MD) simulations. The calculated PMF reveals a global energy minimum, which corresponds to the crystal structure of β2AR-adrenaline complex, and a meta-stable state in which the adrenaline is moved slightly deeper into the binding pocket with a different orientation compared to that in the crystal structure. The orientational and conformational changes in adrenaline during the transition between these two states and the underlying driving forces of this transition are also explored. Based on the clustering of MD configurations and machine learning-based statistical analyses of time series of relevant collective variables, the structures and stabilizing interactions of these two states of the β2AR-adrenaline complex are also investigated.

Identification and characterization of metabolic subtypes of endometrial cancer using systems-level approach

Abstract: Abstract Background Endometrial cancer (EC) is the most common gynaecological cancer worldwide. Understanding the metabolic adaptation and its heterogeneity in tumor tissues may provide new insights and help in cancer diagnosis, prognosis, and treatment. In this study, we investigated metabolic alterations of EC to understand the variations in the metabolism within tumor samples. Methods We integrated the TCGA transcriptomics data of EC (RNA-Seq) with the human genome-scale metabolic model (HMR2.0) and performed unsupervised learning to identify the metabolic subtypes of EC and uncover the underlying dysregulated metabolic pathways and reporter metabolites in each subtype. The relationship between metabolic subtypes and clinical variables was explored. Further, we characterized each subtype at the molecular level and correlated the subtype-specific metabolic changes occurring at the transcriptome level with the genomic alterations. Results EC patients are stratified into two robust metabolic subtypes (cluster-1 and cluster-2) that significantly correlate to patient survival, tumor stages, mutation, and copy number variations. We observed coactivation of pentose phosphate pathway and one-carbon metabolism along with genes involved in controlling estrogen levels in cluster-2, which is linked to poor survival. PNMT and ERBB2 are also upregulated in cluster-2 samples and present in the same chromosome locus 17q12, which is amplified. PTEN and TP53 mutations show mutually exclusive behavior between subtypes and display a difference in survival. Conclusion This work identifies metabolic subtypes with distinct characteristics at the transcriptome and genome levels, highlighting the metabolic heterogeneity within EC.

A Dynamic Time-Bound Access Control for Secure Hierarchical Content Sharing

Abstract: <p>With rapid growth of mobile users, protecting content from unauthorized users become a complex problem. The concept of temporal role-based access control reduces complexity of user management and restricts access to specified time-slots. But, content privacy is still questionable in case of system resources compromise unexpectedly. Therefore, cryptographic solution for time-bound hierarchical content management is an emerging problem. Most of the related schemes focused on individual user keys and/or revocation, but not on time-bound keys. Hence, these are not well suitable for subscription-based services like pay-TV and newspaper. In this paper, we propose a cryptographic time-bound access control with constant size time-bound keys. In our scheme, subscribed time-slots embed into individual user keys to avoid periodical broadcasting of temporal keys. We prove that our scheme is selectively secure against collision attack under chosen-ciphertext attack. We then discuss cloud-based application to show the strategies of efficient revocation and reduced computational overheads for end-users. </p>

F3: Fair and Federated Face Attribute Classification with Heterogeneous Data

Abstract: Fairness across different demographic groups is an essential criterion for face-related tasks, Face Attribute Classification (FAC) being a prominent example. Simultaneously, federated Learning (FL) is gaining traction as a scalable paradigm for distributed training. In FL, client models trained on private datasets get aggregated by a central aggregator. Existing FL approaches require data homogeneity to ensure fairness. However, this assumption is restrictive in real-world settings. E.g., geographically distant or closely associated clients may have heterogeneous data. In this paper, we observe that existing techniques for ensuring fairness are not viable for FL with data heterogeneity. We introduce F3, an FL framework for fair FAC under data heterogeneity. We propose two methodologies in F3, (i) Heuristic-based and (ii) Gradient-based, to improve fairness across demographic groups without requiring data homogeneity assumption. We demonstrate the efficacy of our approaches through empirically observed fairness measures and accuracy guarantees on popular face datasets. Using Mahalanobis distance, we show that F3 obtains a practical balance between accuracy and fairness for FAC. The code is available at: github.com/magnetar-iiith/F3 .