• Artificial Inteligence
• Deep Learning

The contemporary era is marked by the growing significance of artificial intelligence and the classification of data, an ongoing challenge at the heart of technological innovation. The rapid evolution of digital technologies has led to a vast availability of visual data, ranging from personal photographs to extensive datasets used in scientific research. The need to understand, organize, and interpret this enormous volume of visual information has created a pressing demand for AI models capable of addressing the intricate challenges of image classification. This task is fundamental in sectors like medicine, security, and industrial automation. Effectively tackling this challenge requires AI models that dynamically adapt to new data without losing their optimizations and capabilities over time. As the diversity and complexity of images grow, developing flexible and resilient machine learning models is crucial. Image classification represents one of the most relevant challenges in today's AI landscape. Developing models capable of addressing this challenge with flexibility and resilience over time reflects the growing need for intelligent technologies in various aspects of our daily lives. Continuous innovation and advanced research are essential for creating increasingly sophisticated and adaptable AI models to meet the evolving demands of image classification in the digital era. Artificial intelligence is crucial in this experimental thesis project conducted at Vitesco Technologies. AI offers innovative machine-learning techniques for optimizing pre-trained image classification models, aiming to improve accuracy, efficiency, and robustness. Additionally, the thesis explores new frontiers of AI applied to image classification, assessing their potential to enhance existing models significantly.

The thesis project aims to leverage preprocessing techniques, implement and evaluate optimization techniques through controlled experiments, and analyze the results to assess the impact of AI on improving image classification models. This experimental thesis, conducted at Vitesco Technologies, will expand knowledge in this evolving field, providing an in-depth analysis of using AI to optimize image classification models in an industrial context.