ДОСЛІДЖЕННЯ ВИКОРИСТАННЯ ФРАГМЕНТНОГО АНАЛІЗУ ВІДЕОПОТОКІВ ДЛЯ ЗАДАЧІ КЛАСИФІКАЦІЇ

This work is devoted to fragmented video analysis methods as a way to optimize the classification task. Traditional approaches to video classification often require significant hardware resources. The concept of fragmented video stream analysis aims to alleviate these resource demands while enhancing classification effectiveness through algorithmic improvements. By breaking down video frames into smaller components, this method reduces hardware requirements and streamlines the classification process. Incorporating fragmented video analysis not only addresses the resource-intensive nature of traditional classification methods but also opens doors to more nuanced insights into video content. This approach empowers systems to better discern patterns and anomalies, leading to enhanced decision-making processes and user interactions across diverse domains. Additionally, the utilization of unusual matrix metrics further enhances classification efficiency. Overall, this work explores innovative strategies to make video classification more resource-efficient and effective, offering promising avenues for advancing classification techniques in various applications.