Advancements in Computer Vision: Exploring Deep Learning and Transformer-Based Models for Enhanced Visual Perception

Recent advancements in computer vision have significantly transformed various industries, from healthcare to autonomous driving. This paper presents a comprehensive survey of these developments, with a particular focus on deep learning and transformer-based models. We explore the fundamental concepts and methodologies, including feature extraction, classification, segmentation, and object detection. The paper also highlights the evolution of computer vision frameworks and tools, emphasising the contributions of convolutional neural networks (CNNs), generative models, and transfer learning. Additionally, we discuss emerging trends such as vision transformers and multimodallearning, while acknowledging persistent challenges like data scarcity and realtime processing. Through an in-depth analysis, we aim to provide scholars and professionals with a detailed understanding of the current state and future prospects of computer vision. The paper further examines specific applications in healthcare, autonomous cars, retail, agriculture, and security, illustrating how computer vision technologies are redefining established practices and enhancing decision-making capabilities.