Volume 14 | Issue 4 |

 Abstract

"Mahadevi Verma ke sahitya mein Hindi-navjagaran ke sutra"

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

"Mahadevi Verma ke sahitya mein Hindi-navjagaran ke sutra"

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Lip balm is a cosmetic product used to protect and moisturize lips. This project focuses on developing a natural lip balm using beetroot extract as a coloring and antioxidant agent. Beetroot contains betalains, which provide a natural red tint and offer antioxidant benefits. The formulation avoids synthetic dyes and harmful chemicals. The prepared lip balm was evaluated for parameters such as color, texture, pH, spreadability, melting point, and stability. The results showed that beetroot-based lip balm is safe, effective, and eco-friendly.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Beetroot extract, Beta vulgaris, herbal lip balm, natural colorant, betalains, antioxidants, beeswax, coconut oil, shea butter, vitamin E, lip care, stability.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Mango is a significant tropical fruit crop that plays a crucial role in the agricultural economy of India and many other countries. However, mango production is severely affected by diseases such as Anthracnose, Bacterial Canker, and Powdery Mildew, which reduce yield and fruit quality. Traditional manual disease detection methods are time-consuming, subjective, and require expert knowledge, making them inefficient for large-scale farming. To address this challenge, this study proposes an automated mango fruit and leaf disease detection system using deep learning techniques. The proposed model utilizes transfer learning with pretrained convolutional neural networks, namely VGG16 and ResNet, for accurate feature extraction and classification of healthy and diseased samples. Image preprocessing techniques including resizing, normalization, and data augmentation are applied to improve model robustness and generalization. The model is implemented using TensorFlow and Keras, and trained with the Adam optimizer and L2 regularization to reduce overfitting. Performance evaluation using accuracy, precision, recall, and F1-score demonstrates that the proposed deep learning approach outperforms traditional machine learning methods such as SVM, KNN, and Random Forest. The system provides a reliable and scalable solution for early disease detection, contributing to smart and sustainable agriculture.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Mango Disease Detection, Deep Learning, CNN, Transfer Learning, VGG16, ResNet.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Reliable vehicle detection is a cornerstone of autonomous navigation and advanced driver assistance systems (ADAS); how-ever, its efficiency is severely compro-mised under adverse weather conditions, particularly during heavy rain [1], [2]. Rainfall introduces significant visual arti-facts - raindrops on camera lenses, re-duced contrast and light scattering, which resulted to a precipitous drop in the mean Average Precision (mAP) of standard sin-gle-modality detectors [17], [21]. This research proposes a Multi-Modal Fusion framework designed by enhance detecting robustness by integrating complementary data from visible-light cameras and mil-limeter-wave (mmWave) radar [13], [21]. While visible-light cameras provide high-pixel resolution semantic information es-sential for identifications, they are highly capable or admitting to "rain noise" and low-light degradation [3], [11]. In com-parison, mmWave radar maintains opera-tional stability in precipitation because of its longer wavelength, providing accurate ranging and velocity data despite reduced visibility [2], [16]. Our proposed architec-ture utilizes a deep-fusion approach, where feature maps from a YOLO-based detector [8], [22] are enriched with radar-point cloud features at the neck level of the neural network. This permits the mod-el to "attend" to radar-validated spatial regions when optical data is unreliable [17], [22]. The methodology incorporates data aug-mentation techniques--specifically rain-streak simulation and noise injection--to train the model on diverse "wet-weather" scenarios [5], [21].Experimental evalua-tions, conducted using an integration of regional vehicle datasets [6], [15] and synthetic weather-augmented sets, demonstrates that the multi-modal fusion model significantly outperforms single-modality YOLO architectures in higher-intensity rain scenarios [17], [22].Furthermore, the integration of Tem-poral-Spatial attention mechanisms, simi-lar to those used in traffic management [4], ensures that vehicle tracks remain consistent despite intermittent occlusions caused by spray and splash [14], [16]. The results suggest that this fusion strategy provides the necessary redundancy for safety-critical applications in monsoon-prone and mountainous terrains [13], [16].

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Multi-modal fusion [17], Vehicle detec-tion [6, 9, 15, 17], Adverse weather condi-tions [1, 2, 9, 11, 21, 22], YOLO (You On-ly Look Once) [3, 8, 16, 22], Autonomous driving [5, 9, 22], Robustness [1, 17, 21], Deep learning [8, 17, 18], and Real-time systems [4, 16, 19].

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Oblique Wing, Variable Sweep, CFD, Low Reynolds number, UAV, Aerodynamics.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The quick progress of deep learning has resulted in larger and larger models, posing challenges to their computational requirements and environmental implication. Knowledge Distillation (KD) has emerged as an attractive solution to reducing the size of these complex models and the ability to achieve similar levels of performance using smaller and more efficient models. The paper will discuss the recent progress in the knowledge distillation field, focusing on the advancement of the efficiency of algorithms, multiple stages of distillation, and its role in Green AI efforts. It discusses the background research and recent advances, especially when it comes to transformer-based architectures and TinyML applications. The results show that knowledge distillation is an effective method to deploy high-performance models on resource-limited edge devices, as well as minimize energy usage and carbon emissions.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Knowledge Distillation, TinyML, Model Compression, Green AI, Neural Networks, Efficiency

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Intranasal drug delivery has emerged as a promising non-invasive strategy for direct brain targeting by bypassing the blood-brain barrier via the olfactory and trigeminal nerve pathways. However, conventional intranasal formulations suffer from limited nasal residence time and poor mucosal absorption. Recent advances in nanotechnology have enabled the development of diverse nanocarrier systems, including polymeric nanoparticles, lipid-based carriers, micelles, dendrimers, and nanoemulsions, which significantly enhance drug solubility, stability, permeability, and brain bioavailability. Anatomical and physiological features of the nasal cavity play a crucial role in determining delivery efficiency, while biocorona formation influences nanoparticle pharmacokinetics and biodistribution. Stimuli-responsive and targeted nanocarriers offer improved spatiotemporal control of drug release and cellular uptake. Despite encouraging preclinical outcomes, challenges related to toxicity, scalability, and regulatory approval hinder clinical translation. This review summarizes recent advancements in nanotechnology-based intranasal delivery systems and highlights their potential in improving therapeutic outcomes for central nervous system disorders.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Nanotechnology, Mucoadhesion, nose-to-brain, nanoparticles, micelles, emulsions, liposomes, SLNs, NLCs, blood-brain barrier, bioavailability.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Human-computer interaction has also changed and taken a new direction whereby voice command systems have been adopted instead of the traditional type of tactile inputs. This change has brought down the intellectual and physical burden on customers, which facilitates smooth digital processes. In this paper, the development of J.A.R.V.I.S. is described. Another project, (Just A Rather Very Intelligent System), is a desktop based virtual assistant developed on the Python programming ecosystem. In contrast to commercial solutions that are using clouds to process information and require users to open accounts, J.A.R.V.I.S. emphasizes on local processing and user privacy. The system makes use of SpeechRecognition library to process acoustic signals and pyttsx3 to do speech synthesis offline. The user experience is improved by a well-built graphical interface that is created using the PyQt5 library. The paper will analyze the architectural design, algorithmic code as well as functional units such as email automation through SMTP and media automation through pywhatkit. According to the performance analysis, the system can be used as a stable, responsive and privacy-sensitive alternative in the automation of routine desktop tasks. Index Terms--Artificial Intelligence, Desktop Automation, Human-Computer Interaction, Python, Speech Recognition, PyQt5, Virtual Assistant

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Computer Science

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The growing use of digital technology in education has considerably increased the usage of online learning platforms, resulting in higher amounts of daily screen exposure. While digital learning is convenient and flexible, concerns have been raised about its possible impact on cognitive functioning. The current study sought to investigate the association between screen time and cognitive failures among online learners, as well as the function of metacognitive awareness in that relationship. Specifically, the study looked into whether metacognitive awareness mediates the link between screen time and ordinary cognitive failures. A quantitative correlational research methodology was used, with data collected from 147 online learners via standardized self-report measures assessing screen time, metacognitive awareness, and cognitive failures. Statistical descriptive measures and analyses -- including correlation tests and mediation analysis -- were used to explore how the variables interacted together. Reviewing the findings as a whole, it was determined that there is a direct correlation between screen time and cognitive impairments. In other words, if people spend more time using screens, their probability of experiencing cognitive failures will increase. In addition, screen time has a positive correlation to metacognitive awareness. Thus, it can be reasonably assumed that using screens to engage in learning (through digital learning environments) will improve participants' metacognitive awareness. However, there is no statistically significant correlation between metacognitive awareness and cognitive impairment. Mediation Analysis has indicated that an awareness of metacognitive processes alone does not eliminate the relationship between the amount of time spent on screens and cognitive performance. Therefore, while learners may understand how they learn, this understanding cannot negate the negative consequences of excessive screen use on cognitive performance. In conclusion, this study adds to the existing literature examining the effects of screen use on cognitive functioning and investigating the relationship between cognitive failures and screen use. Hence, in order to avoid cognitive interference while using the internet for learning, students must learn to develop skills of good screen use along with metacognitive skills.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Screen time, cognitive failures, metacognitive awareness, online learners, digital learning, cognitive functioning.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Menopause is a significant developmental transition characterized by complex biological, psychological, and social changes. The present study examines mental and physical health during perimenopause and menopause using a biopsychosocial framework. It highlights the role of hormonal fluctuations, psychosocial stressors, and cultural influences in shaping women's experiences. The study emphasizes the importance of awareness, early identification of symptoms, and holistic interventions to improve quality of life.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Menopause; Perimenopause; Mental Health; Depression; Anxiety; Stress; Menopause Rating Scale (MRS); DASS-21; Hormonal Changes; Vasomotor Symptoms; Sleep Disturbances; Biopsychosocial Model; Women's Health; Quality of Life 1. Introd

  License

Creative Commons Attribution 4.0 and The Open Definition