Journal of Advanced Research in Computer Graphics and Multimedia Technology

Impact of Social Media on Teenagers' Health

Vaishnavi Suryakant Nemade, Rajashri Sanjay Koli, Varsha Ravindra Chaudhari — Wed, 30 Sep 2026 00:00:00 +0000

Social media has become an important part of adolescent life, providing opportunities for communication, learning, entertainment, identity development, and social support. At the same time, frequent or problematic use may be associated with sleep disruption, anxiety, depressive symptoms, cyberbullying, body-image concerns, unhealthy eating, reduced physical activity, and exposure to health-risk behaviors. This paper reviews research on the impact of social media on teenagers' health, with particular attention to mental health, sleep, physical health, social relationships, and risky behaviors. Evidence from systematic reviews, meta-analyses, and public health guidance indicates that the relationship is complex rather than universally harmful: outcomes depend on the type of content, manner of use, duration, individual vulnerability, social context, and platform design. Recent evidence also suggests that problematic use is more consistently associated with adverse outcomes than simple frequency of use. The paper discusses positive and negative effects, major risk factors, preventive strategies, and research gaps. It concludes that balanced use, digital literacy, supportive family and school environments, safer platform design, and continued research are needed to protect adolescent health while retaining the benefits of online social connection

Eye Tracking Based Virtual Mouse Using Webcam and Computer Vision

Wed, 30 Sep 2026 00:00:00 +0000

Most people still interact with a computer through a mouse, a keyboard, or a touchpad, none of which are usable by everyone. This project explores whether an ordinary webcam, without any specialised eye-tracking hardware, can be turned into a working substitute for a mouse. The prototype is built in Python around OpenCV for video capture, MediaPipe for facial and IRI landmark detection, NumPy for the underlying geometry, PyAutoGUI for using mouse commands on the operating system, and Tkinter for a small control panel. A single RGB camera watches the user's face; the position of the iris within each eye is estimated frame by frame, and a nine-point calibration routine learns how that position maps onto the screen. To keep the cursor from jittering, its motion is smoothed rather than updated instantaneously. Left and right clicks are triggered by closing one eye at a time, while a longer, deliberate closure of both eyes switches the system into a text-selection mode; looking toward the top or bottom of the frame scrolls the page. The system should be read as a working proof of concept rather than a finished product — it has not yet been through a formal user study, so this paper reports what was built and how and leaves the accuracy number and completion-time figure for the evaluation described in Section VIII.

Skin Cancer Detection through Image Analysis with a Dual- Architecture Deep Learning Approach

Swetta Kukreja, Deepa Parasar, Tejas Dashpute, Rohan Gaikwad, Suyash Dubey — Thu, 14 May 2026 00:00:00 +0000

This research paper addresses global skin cancer concerns, emphasising the need for early detection through artificial intelligence (AI), specifically leveraging AlexNet and EfficientNet. The study highlights limitations in traditional diagnostic methods and proposes an AI-driven paradigm shift. In this paper details of the development of a deep learning-based approach have been presented, wherein existing differentiating images are augmented, covering data pre-processing, training, and evaluation with rigorous scientific methodology that utilises the parametric classification of a skin cancer-based lesion or visible marking. Findings emphasise the efficacy of both models, envisioning a future where AI and image-based observation play pivotal roles in early skin cancer detection. The AlexNet architecture receives a 98.9% accuracy, while the EfficientNet B1 provides an accuracy of 88.9%. The conclusion underscores the transformative potential of AI and a productive way of combining architecture to multiply the efficacy in skin cancer detection, leading the way for increased accuracy and accessibility in early diagnosis.

DOI: https://doi.org/10.24321/3051.4266.202601

Analysis of FCN8 and YOLOv8 in Enhancing Road Segmentation for Autonomous Driving

Thu, 14 May 2026 00:00:00 +0000

In the rapidly advancing field of autonomous driving, achieving precise road segmentation is paramount for accurate environmental perception. This research paper introduces an innovative road segmentation approach, which leverages the FCN-8 architecture with VGG16 as its backbone, further enhanced through the implementation of the sigmoid activation function and the Adam optimiser. In addition, we conduct a comprehensive comparative analysis with the YOLOv8 architecture to provide a holistic view of road segmentation solutions in this context. This research is based on a custom dataset specifically curated to address the unique challenges of autonomous driving scenarios, ensuring that the models are tested against real-world conditions. Our evaluation demonstrates the robustness and efficiency of both approaches in the context of autonomous driving, shedding light on their respective strengths and weaknesses. This innovative methodology not only advances road segmentation accuracy but also significantly contributes to the realm of self-driving technology, fostering safer and more reliable autonomous vehicles in real-world applications.

DOI: https://doi.org/10.24321/3051.4266.202602

Integration of Artificial Intelligence with AR/VR

Ubaid Khatik, Mohd Faizan Khan, Vivek Ghate, Rutvik S Patil — Wed, 30 Sep 2026 00:00:00 +0000

Artificial Intelligence (AI) is becoming an important technology for improving the capabilities of Augmented Reality (AR) and Virtual Reality (VR) systems. AR and VR provide immersive and interactive environments, while AI can make these environments more intelligent, adaptive, and responsive to user actions. This research explores the integration of artificial intelligence with AR/VR technologies to improve user interaction, personalization, and the overall immersive experience. The study focuses on the use of AI techniques such as computer vision, machine learning, natural language processing, and speech recognition in AR/VR applications. AI can help AR/VR systems understand real-world objects, recognize user commands, analyze user behavior, and provide suitable responses in real time. This integration has potential applications in education, healthcare, gaming, industrial training, manufacturing, and other interactive environments. The research also examines the major benefits and challenges of AI-enabled AR/VR systems, including computational requirements, privacy, system accuracy, hardware limitations, and user safety. The study aims to provide an overview of how AI can enhance existing AR/VR technologies and contribute to the development of more intelligent and personalized immersive applications. By combining intelligent perception with immersive visualization, AI can also support context-aware assistance, improve accessibility, and enable more effective decision-making during complex virtual and augmented tasks.

Deepfakes and Digital Trust: The New Challenge for Cybersecurity

Wed, 30 Sep 2026 00:00:00 +0000

Deepfake technology represents a rapidly evolving application of artificial intelligence that enables the creation of highly realistic synthetic media. The increasing accessibility of deepfake-generation tools has created new opportunities for cybercriminals to conduct sophisticated attacks against individuals, businesses, and public institutions. Unlike conventional cyber threats, deepfake attacks exploit not only technical vulnerabilities but also human perception and trust. This research investigates how deepfakes influence digital trust and create new challenges for cybersecurity. The study focuses on major threats, including identity manipulation, social engineering, business email compromise, financial fraud, misinformation, and biometric authentication attacks. It also evaluates the effectiveness and limitations of current deepfake detection approaches, recognizing that the rapid development of generative AI can enable attackers to overcome existing detection mechanisms. The research proposes that effective protection requires a layered security framework involving multi-factor authentication, human verification, AI-assisted detection, digital content provenance, employee awareness, and continuous security monitoring. The findings emphasize that maintaining digital trust in an AI-driven environment requires organizations to reconsider traditional assumptions about the authenticity of digital communication and develop adaptive cybersecurity strategies capable of addressing emerging synthetic-media threats.

Transforming Women’s Empowerment through Emerging Technologies: The Role of Augmented and Virtual Reality in Digital Inclusion

Nikita Warke, Shubhangi Sapkale, Nayana Chaudhari, Punam Saraf, Shubham Adawal — Wed, 30 Sep 2026 00:00:00 +0000

The fast development of Information Technology (IT) is influencing the way of women's access to education, employment, entrepreneurship, healthcare-related information, financial services, and social interaction. However, the digital revolution does not apply evenly across the globe. In many countries, women and girls face difficulties connected with access to internet, devices, digital skills, norms, safety, and unequal participation. AR and VR are offering new opportunities, because these technologies have the capability to create immersive, visual and practice-based digital environment. This paper looks at how AR and VR can empower women through their digital inclusion, paying specific attention to education, digital and professional skills development, employability, entrepreneurship, STEM exposure,

healthcare-related learning, confidence-building and social participation. The study applies a secondary, conceptual and analytical methodology based on the literature review and evidence provided by international organizations. The analysis reveals a great potential of immersive technologies for experiential learning, job practice, training, remote teaching, product visualization, networking, and confidence building. At the same time, the study admits that advanced technologies will reproduce inequality without basic digital access. Main concerns are related to affordability, connectivity, digital literacy, accessibility, privacy, cybersecurity, technology-facilitated violence, and absence of gender-sensitive content. Therefore, the paper suggests a staged inclusion strategy starting from connectivity and digital literacy and following affordable devices, AR/VR exposure, skills development, and economic opportunities. The conclusion is that AR and VR should be incorporated in broader women's digital inclusion programs rather than used as standalone tools.

Deep Learning–Based Detection of Deep Fakes Using Convolutional and Recurrent Neural Networks

Shital Jadhav , Sonal Patil — Wed, 30 Sep 2026 00:00:00 +0000

The exponential growth in the generation and dissemination of deep fake content has resulted in huge challenges to the integrity and trustworthiness of digital media. The current paper tries to meet an unmet need for robust and reliable deep fake detection techniques by reviewing the current literature and meta-analyzing the existing methods. This study thoroughly assesses various deepfake detection models, their methodologies, strengths, and weaknesses, and performance metrics, such as accuracy, precision, recall, and F1-scores. The paper provides a review and critical analysis of methods using a diverse set of techniques, ranging from CNNs, RNNs, GANs, and hybrid approaches. CNNs, especially those using transfer learning from pre-trained models like VGG16, ResNet, and InceptionV3, have demonstrated image-based deep fake detection with high accuracy because of their capability in effectively capturing spatial features. RNNs, especially the Long Short-Term Memory (LSTM) architecture, are applied to perform video-based detection, leveraging the distinctive features of analyzing data over a temporal sequence and identifying inconsistencies over time. In the beginning, GAN-based methods used to generate deep fakes have now been repurposed for detection. This work provides a technique to repurpose GAN-based methods for detection using the capability of such models to distinguish between real and synthetic data using adversarial training. Hybrid architectures that integrate several such neural network models hold promise for improving detection performance through combined strengths. This comparative analysis demonstrates that CNN-based methods are typically the best-performing methods for static image analysis, while RNN is more effective for the video sequence. GAN-based methods are capable of adapting to new deepfake generation techniques but do so with increased computational complexity. Hybrid methods have increased complexity but present a better trade-off between accuracy and computational efficiency. The impact of such a detailed review is two-fold. By systematically categorizing and reviewing the existing methods for deep fake detection, the current study brings out a clear view of the state-of-the-art, guiding researchers and practitioners in the selection of relevant techniques with a view to specific application requirements.

Further, the identification of limitations and gaps in the currently existing methods lays out research directions that contribute to enhancing the security and reliability of digital media, fostering greater public trust in digital content authenticity.

A Comprehensive Review on Machine Learning Techniques for Oral Cancer Diagnosis and Prognosis

Saima Suraiya, Arun Kumar Rai — Sun, 21 Jun 2026 00:00:00 +0000

Oral cancer persists as an unsolved global health dilemma with a very low survival rate and is begging for early diagnosis and prognosis. Machine learning systems and data mining techniques have revolutionised conventional techniques for diagnosing these cancers by facilitating automated, accurate, and fast detection of oral lesion differentiation. This paper asserts the application of a broad evaluation of existing techniques, which include machine learning algorithms in predicting oral cancer, for example, Support Vector Machines (SVM), K-Nearest Neighbours (KNN), Logistic Regression, Random Forest, Convolutional Neural Networks (CNN), Neural Networks (NN), and deep learning approaches. The frameworks have demonstrated much potential in accurately creating contrasts from histopathological images and clinical data and categorising benign against malignant forms. The models that are compared showed that in cases, especially where feature selection and dimensionality reduction techniques were enforced, neural networks and convolutional neural networks (CNNs) performed significantly better than linear and statistical machine learning methods. The addition of data preprocessing was found to increasingly improve the model performance. Methods like normalisation, augmentation, and feature extraction are crucial for model robustness and generalisation. The review paper also brings forth emphasis on achieving better predictive performance with combined learning and ensemble approaches. Major challenges still persist because of skewed data distributions, non-interpretability, and the combinatorial limitations of multimodal integration. Inclusion of clinical, genetic, and radiographic data apparently promises an obvious solution with explainable AI advancements supporting clinical decision-making. The article finally points at an untapped potential in the field, showing infinite promise for innovative technology for early detection, diagnosis, and efficient treatment – leading to increased survival.

How to cite this article:
Suraiya S, Rai A K. A Comprehensive Review on Machine Learning Techniques for Oral Cancer Diagnosis and Prognosis. J Adv Res Comp Graph Multim Tech. 2026; 8(2): 13-22.