Volume 13 | Issue 8 |

 Abstract

Machine learning algorithms developed on large- scale audio datasets have helped AI-generated music make a lot of progress in the last few years. This paper talks about a music generation system that works for a specific genre and improves Meta's MusicGen model by using the Free Music Archive (FMA) dataset. A mobile app built with React Native lets users choose a genre and create, play, share, or download music interactively. The research discusses about the architecture, data preprocessing techniques, model fine-tuning strategies, integration with mobile apps, and changes made to address challenges with the base MusicGen model.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Music Generation, Generative AI, MusicGen Model, Deep Learning, Neural Networks, React Native, Mobile Application, Fine-tuning, Audio Synthesis, Conditional GANs, Music Information Retrieval, Artificial Intelligence, Audio Pro- cessing, Genre-specific Music, Transfer Learning.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

ABSTRACT -- We present a technique for identifying surface flaws in railroad tracks using the YOLOv8 model, which is designed to overcome the challenges of detecting small and hidden targets. To improve the model's attention mechanism, we replace the original folding of YOLOv8n with the SPDCONV block while maintaining the backbone network and the overall architecture of the original model. We also integrate the EMA Atones mechanism into the neck component, enabling the model to leverage data from various attributes and enhancing its feature representation capabilities. Additionally, we replace the original CIOU loss function with YOLOv8's Focus SIOU loss function, which adjusts the weights of positive and negative samples to better penalize difficult examples. This modification significantly enhances the model's ability to detect challenging instances, ensuring that each target receives more focused attention from the network. As a result, the model's overall performance and effectiveness are improved. Experimental results provide compelling evidence of the improved algorithm's enhanced accuracy and recall. Compared to the original YOLOv8n model, the extended version achieves average accuracy, recall, and precision rates of 93.9%, 93.7%, and 91.1%, respectively, reflecting improvements of 3.6%, 5.0%, and 5.7% . Notably, these gains are achieved without increasing the number of parameters or the size of the model. The improved method demonstrates high effectiveness in detecting surface flaws on railroad tracks.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

A. Station module: The Station Module is engineered to track platform availability via infrared sensors, facilitating effective train scheduling and routing. Future enhancements could involve the integration of cloud-based analytics and AI- powered dispatch systems, enabling real-time platform assignments based on train locations, traffic levels, and delay forecasts. Furthermore, linking the Station Module to a centralized control network would allow for synchronized management across various st

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This study investigates the influence of metakaolin (MK) and colloidal silica (CS) as partial replacements for cement on the mechanical properties of M60 grade concrete. Four mix proportions with varying MK (10% to 30%) and CS (0% to 6%) were evaluated for compressive strength, split tensile strength, and flexural strength at curing periods of 7, 14, and 28 days. Results indicate that the mix containing 10% metakaolin and 2% colloidal silica (MD2) exhibited the highest compressive strength, achieving up to 37% greater strength than the conventional concrete at 28 days. Split tensile and flexural strength tests showed improved performance with lower percentages of colloidal silica, with the optimal tensile strength also observed in the MD2 mix. Higher replacement levels (20%-30% MK and 4%-6% CS) led to a decrease in tensile and flexural strengths and did not consistently improve compressive strength beyond early curing ages. The findings suggest that 10% metakaolin combined with 2% colloidal silica is the most effective replacement for enhancing the overall mechanical properties of M60 grade concrete.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Metakaolin, Colloidal Silica, M60 Grade Concrete, Compressive Strength, Split Tensile Strength, Flexural Strength.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Abstract Abstract : aper is focused on the precarious circumstances of the Particularly Vulnerable Tribal Groups through out the country with special reference to the PVTGs of Andaman and Nico bar Islands. For better understanding, a brief description is given on the criteria on the basis of which some of the tribes are identified under this category. The PVTG is the concept substituted by primitive tribe, a sub-group among the Scheduled Tribes was identified during the fifth plan owing to their backwardness. In 1973, the Dhebar Commission introduced the concept to prioritise their protection and development as development has been slow among them and some of them are still at the stage of hunting and food gathering and yet to cultivate the land. There are 75 PVTGs who inhabit in Northern, North-Eastern, Eastern, South-Western, Southern and middle regions of India. However, thirteen types of PVTGs are found in Odisha alone which is highest among the states and Union Territories. There are twelve PVTGs whose population is more than 50,000. The Sahariya of Rajasthan and the Kutia-Kandha of Odisha have a sizable population. However, some of them show a declining trend of population and have become engendered human groups like the Great Andamanese, the Onge, the Jarwa and the Shompen of Andaman and Nico bar islands and the Toto of Jalpaigudi in West Bengal. It is anticipated that one day they may extinct. All of them belong to the PVTGs except the Nicobarese whose population is above 50,000 and it is hig... On the other hand, all other communities are found to be vulnerable according to the criteria fixed by the ministry. Among many factors, the British interference into their life has made irreparable damage as a result of which a kind of infection has spread among them. The disease seems to be incurable due to which there has not been population growth among them. The public interference, loss of autonomy to their land, the exchange of food items with outsiders have opened many ways of exploitation of the isolated groups. The Great Andamanese, the Onge as being friendly have become the pray of the conspiracy of the so called civilized people. On the other hand, the Sentinelese being the most isolated tribe who never interact with anyone not even with the other tribal groups in the island. They never allow anyone to enter into their residential area and remain independent of foreign relations. Therefore, have comparatively more number of population. Owing to their vulnerability, the Andaman administration has taken all the responsibilities of by providing free ration, medical care, etc. However, there is no option left for them to be recovered. The philanthropists, social workers, researchers, NGOs are urged to collaborate with the Government to rehabilitate the engendered PVTGs of the island.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Particularly Vulenrable British Interference Consult Consequences Infectious Disease, Engendered Group,Irreparable Harm,Declining Population

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The positivist paradigm and the quantitative research technique formed the basis of this study's research design. A total of 421 respondents from Haryana's educational institutions provided the data. Using structural equation modelling (SEM), the study's model was evaluated. The study's findings suggested that the individual's attitude is influenced by performance expectancy, perceived service quality, facilitating conditions, and effort expectancy. The intention to recommend is influenced by the behaviour intention to use, which is in turn influenced by the attitude.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

E-government, e-gov, UTAUT model, TAM model, UMEGA model

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Recommender systems have become pivotal in the digital era, enhancing customer experiences by providing personalized suggestions across different domains such as e commerce, entertainment, and social media. This paper presents a comprebensive literature survey and proposes a novel hybrid framework that integrates image based data with advanced mathemnatical models improve recommendation accuracy and user engagement. We develop an image enhanced matrix factorization method and a visually aware neural collaborative fltering model, and we evaluate both on standard benchmarks, achieving a 15 % improvement in F1 score over leading baselines. Detailed mathematical formulations of each algorithm are provided, along with a thorough comparison of efficiency and scal ealability, The results demonstrate that ncopor visual features significantly enbances the relevance and recommendations, ultimately boosting customer satisfaction.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Recommender Systems, Customer Experience, Image Processing, Machine Learning, Customer Behavior, Hybrid Approach I.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

When the English conquered the world, the English language slowly spread around the globe. It did not remain just the language of the small island of England. In course of time, the language being widely spoken came to be known as a global language in today's world. The English language has several loanwords. The language not just borrowed from other European languages, but also from different Indian languages. Old English or the language of the Celts is very different from modern English. When the nomadic tribes dwelled in the island, they went through various invasions and influences, thus slowly affecting the language. The French, Scandinavian and Latin influences made it what it is today. The language Sanskrit, being a very important part of Indian art and culture, has also greatly influenced English. Many words of common use including several names have been taken from the language Sanskrit. English, being the official language of many countries, has a large number of speakers. It is not just spoken in the United States of America and England, but also in Afro-Asian countries. Even non-English speaking countries encourage children to take up English in School as a second language. Being widely used today, English has now become a language of science and technology.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

English Language, Communication, Sanskrit

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Abstract This article critically examines V.S. Naipaul's India trilogy An Area of Darkness, India: A Wounded Civilization, and India: A Million Mutinies Now as exemplary postcolonial travel writing that navigates diasporic consciousness and identity formation. It argues that Naipaul's positionality as a Trinidadian of Indian descent educated in Britain situates him as both insider and outsider, allowing him to produce a complex critique of India's colonial legacies, cultural hybridity, and social transformation. Through close reading and intertextual analysis, this article demonstrates how Naipaul's ambivalence functions methodologically, revealing the tensions between longing, critique, and recognition of pluralism in postcolonial India.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Keywords: V.S. Naipaul, Postcolonial Travel Writing, Diaspora, India Trilogy, Cultural Hybridity, National Identity, Colonial Legacy

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

An object detection processor based on a binarized neural network (BNN) architecture is developed to enable efficient deployment on resource-constrained FPGA platforms. Traditional convolutional neural networks (CNNs), while highly effective for visual recognition tasks, impose significant demands on computational resources and memory bandwidth, present challenges for embedded applications. To address these limitations, the proposed system leverages binary operations, on-chip memory usage, and an optimized neural architecture to minimize complexity and power consumption. The design achieves accurate object detection performance within the limitations of FPGA hardware, highlighting its potential for efficient inference in real-time environments. Its modularity and scalability allow for adaptation across a wide range of embedded vision scenarios, particularly in domains requiring low latency and high energy efficiency. The proposed approach demonstrates the feasibility of integrating deep learning models into compact, resource-effective hardware systems, offering a promising direction for future applications in areas such as IoT monitoring, smart sensors, and mobile robotics.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Object Detection, FPGA, Binarized Neural Network, Embedded Vision, Resource Optimization, Low-Power Design, IoT, Robotics, Hardware Acceleration.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The domain of medical image processing focuses on the early detection of lung cancer using advanced image processing techniques. Currently, lung cancer diagnosis primarily depends on manual analysis of chest X-rays or CT scans by radiologists. While experienced professionals can identify abnormalities, the process is time-consuming, prone to human error, and often limited by subtle signs that are difficult to interpret in early stages. Existing computer-aided detection systems are available, but many suffer from low accuracy, high false-positive rates, and limited generalization to diverse datasets. The system is designed to process standard datasets of lung images, performing key steps such as image preprocessing to enhance quality, feature extraction to identify relevant patterns, and classification to accurately distinguish between cancerous and non-cancerous images. The dual-phase approach consists of a training phase, where the CNN model learns from labeled data, and a testing phase, where the model is validated with unseen images to measure its performance. The application of CNNs allows the model to learn complex features that are difficult to capture through traditional handcrafted techniques, leading to improved accuracy and reduced false positives. Overall, integrating advanced image processing techniques with machine learning models such as CNN offers promising improvements in lung cancer diagnosis. This approach not only accelerates the diagnostic workflow but also supports healthcare professionals by providing consistent and reliable analysis of medical images. The proposed system has the potential to improve patient outcomes by facilitating timely detection and treatment, ultimately contributing to more effective management of lung cancer on a broader scale.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Lung Cancer Detection, CT Imaging, Image Processing, CNN, Medical Diagnosis

  License

Creative Commons Attribution 4.0 and The Open Definition