Real-time detection of violent behavior through surveillance technologies is increasingly important for public safety. This study tackles the challenge of automatically distinguishing violent from non-violent activities in continuous video streams. Traditional surveillance depends on human monitoring, which is time-consuming and error-prone, highlighting the need for intelligent systems that detect abnormal behaviors accurately with low computational cost. A key difficulty lies in the ambiguity of defining violent actions and the reliance on large annotated datasets, which are costly to produce. Many existing approaches also demand high computational resources, limiting real-time deployment... More >

The Generalized Intersection over Union (GIoU) and the Manhattan distance between axis-aligned boxes represented either as corner coordinates or their center and size, are extended to accept a range of bounding boxes as ground truth, producing the metrics RIoU, $R_1$ and $R^t_1$, respectively. In the context of Table Detection it is shown that this box relaxation procedure allows training object detection models with partial or inexact annotations. For the Table Structure Recognition task, several code improvements to Microsoft's open-source Table Transformer increase all $\mathrm{GriTS}$ metrics on PubTables-1M, with the overall accuracy increasing from 0.8326 to 0.8433. Then box relaxation... More >

Lung diseases such as COVID-19, pneumonia, and tuberculosis remain major public health challenges worldwide, emphasizing the urgent demand for accurate and efficient diagnostic methods. This research explores the use of a Convolutional Neural Network (CNN)-based framework for binary classification of chest X-ray images to detect abnormalities. The methodology incorporates preprocessing techniques such as image resizing, normalization, data augmentation, and grayscale transformation to improve input data quality. CNN architecture comprising convolutional, pooling, fully connected, and dropout layers were trained and evaluated on publicly available datasets. The model attained a test accuracy... More >

The rapid advancement in machine learning and artificial intelligence has significantly enhanced capabilities in multimedia content creation, particularly in the domain of deepfake generation. Deepfakes leverage complex neural networks to create hyper-realistic manipulated audio-visual content, raising profound ethical, societal, and security concerns. This paper presents a comprehensive survey of contemporary trends in deepfake video research, focusing on both generation and detection methodologies. The study categorizes deepfakes into three primary types: facial manipulation, lip-synchronization, and audio deepfakes, further subdividing them into face swapping, face generation, attribute m... More >

The challenge with multifocus images lies in different regions being in focus across various shots, resulting in some areas appearing blurry while others are sharp. This issue is prevalent in fields such as medical imaging, remote sensing, and photography, where clear and detailed images are essential. This project introduces a novel approach to multifocus image fusion by integrating the Marr--Hildreth edge detection technique with Discrete Cosine Transform (DCT), Stationary Wavelet Transform (SWT), and Discrete Wavelet Transform (DWT). The Marr--Hildreth algorithm detects edges by identifying zero-crossings in the Laplacian of a Gaussian-blurred image, effectively highlighting areas with si... More >

Deep learning models are pivotal in the advancements of Artificial Intelligence (AI) due to rapid learning and decision-making across various fields such as healthcare, finance, and technology. However, a harmful utilization of deep learning models poses a threat to public welfare, national security, and confidentiality. One such example is Deepfakes, which creates and modifies audiovisual data that humans cannot tell apart from the real ones. Due to the progression of deep learning models that produce manipulated data, accurately detecting and classifying deepfake data becomes a challenge. This paper presents a groundbreaking IRV2-Hardswish Framework for deepfake detection, leveraging a hyb... More >

Plant diseases create one of the most serious risks to the world's food supply, reducing agricultural production and endangering millions of people's lives. These illnesses can destroy crops, disrupt food supply networks, and increase the danger of food deficiency, emphasizing the importance of establishing strong methods to protect the world's food sources. The approaches of deep learning have transformed the field of plant disease diagnosis, providing sophisticated and perfect solutions for early detection and management. However, a prevalent concern with deep learning models is their susceptibility to a lack of generalization and robustness when faced with novel crop and disease categorie... More >

Image fusion, especially in the context of multi-focus image fusion, plays a crucial role in digital image processing by enhancing the clarity and detail of visual content through the combination of multiple source images. Traditional spatial domain methods often suffer from issues like spectral distortion and low contrast, which has led researchers to explore techniques in the frequency domain, such as the Discrete Cosine Transform (DCT). DCT-based methods are particularly valued for their computational efficiency, making them a strong alternative, especially in applications like image compression and fusion. This study focuses on DCT-based approaches, including variants that incorporate Si... More >