Anurag Engineering College

With high prevalence rates for mental health disorders and the emerging interest in digital behavioural data, computational approaches for depression detection have attracted increasing research attention in recent years. Works available in the literature primarily follow a single-modality approach, using either an audio, text, or visual signal, resulting in inadequate reflection of heterogeneous depressive features and the mitigation of high-dimensional behavioural signals. However, any such approaches typically struggle with the following issues, limiting their practical deployment: modality imbalance, robustness to heterogeneous data sources, and transparency. To overcome such limitations, we present a novel attention-guided multi-source multimodal depression detection framework, named MoodSenseAI, that constructs an ensemble deep learning model from multimodal data, including text, speech, and video. Our framework is based on a DeepMoodNet model with multiple modality-specific encoders (TextMoodEncoder, SpeechMoodEncoder, and FaceMoodEncoder) that learn rich semantic, acoustic, and behavioural representations from independent benchmark datasets for each modality. A modality-specific attention-based fusion module aggregates these embeddings at the representation level, thereby providing context-dependent importance weights that enable adaptive modelling of complementary depression-related patterns across heterogeneous datasets. The experimental results on standard benchmark datasets show that the proposed framework achieves significantly higher accuracy, precision, recall, and AUC-ROC than unimodal baselines and improper fusion strategies (macro accuracy = 94.3

This study introduces a distributed deep learning framework for scalable analytics in heterogeneous large-scale environments. The proposed architecture combines data preprocessing, mutual-information-driven feature optimization, distributed neural training, and adaptive learning within a unified analytical pipeline. Experimental validation was performed using benchmark large-scale datasets in distributed computing environments. The obtained results demonstrate notable improvements in predictive accuracy, computational efficiency, and scalability when compared with conventional centralized approaches. The proposed framework achieved an accuracy improvement of approximately 7% while reducing execution time by nearly 32%, highlighting its suitability for real-time large-scale data analytics applications.

This work examined the development of new molecular interactions in liquid mixtures of pyridine with dialkylamines (diethylamine, dipropylamine, and dibutylamine). In order to do this, measurements of the density and speed of sound of these liquids and their binary mixtures have been made at atmospheric pressure and temperatures 298.15, 303.15, and 308.15 K. Using observed data, thermodynamic properties such as excess molar volume, apparent molar volume, isentropic compressibility, and excess molar isentropic compressibility were determined. Further, the measured speed of sound was compared to the speed of sound using Jacobson’s free length theory and Schaff’s collision factor theory. Moreover, the Redlich–Kister polynomial is used to verify the computed excess molar volume and excess molar isentropic compressibility. The V_m^E and k_s,m^E for the examined binary liquid systems deviate negatively from the ideal axis. Additionally, the thermophysical parameters data for the systems indicated above were analyzed using the Jouyban–Acree (J–A) model.

Hiding information in a carrier is called Steganography, whereas the process of detecting the presence of hidden information within any carrier such as digital images, audio, text is known as Steganalysis. In the recent decade the advent of Convolutional Neural Networks(CNN) has boosted the performance of Steganalysis techniques to higher level and eliminated the feature selection phase which requires considerable domain expertise. But in spite of better performance than traditional machine learning based Steganalysis, CNN based techniques have disadvantages like, huge number of parameters and moreover their performance cannot be interpreted. These are major problems in the case of limited data and subtle stego noise. In this paper we discuss and assess the applicability of Kolmogorov-Arnold Networks(KANs) for the task of Steganalysis instead of Convolutional Neural Networks(CNN). KAN networks use learnable spline functions instead of fixed point wise activation used in CNNs. KANs also offer higher interpretability The KolmogorovArnold Networks architecture consists of high-pass aware spline activations, adaptive channel gating mechanism and curvature regularization. As a result the number of parameters decreases compared to CNN. BOSSBase 1.01 dataset is used in our work and the detection accuracy of KAN based Steganalysis is found to be on par with the contemporary CNN based Steganalysis techniques such as SRNet and Yedroudj-Net with requirement of 40% less parameters. This spline based representation leads to interpretable activation patterns with which we can have the insight into the decision process of the Steganalysis model. These results establish the potential of functional neural networks as an efficient and transparent alternative for modern image Steganalysis.

The real-time tracking as well as patients’ surveillance is pervasive in a healthcare domain as the outcome of Internet of Things (IoT) and cloud. The incorporation of IoT with the conventional healthcare systems has enhanced the quality of services (QoS) in healthcare. However, the significant problem in which it cannot be eliminated in such incorporation is the data security which is transferred among client and server. Hence, this research proposes the novel security framework through the combination of blockchain technology with the optimized blowfish algorithm (OBA). The blockchain is utilized and exchanged to secure the integrity of the data and developed the individual device determination with the low power consumption and provides significant effectiveness. The OBA approach is utilized to enhance the confidentiality of the data while the transmission of the data to the server. The results show that the proposed OBA approach attains the accuracy of 97.84