A Novel Approach for Intermittent Video Background Reconstruction

This manuscript presents a novel approach for reconstructing the backgrounds of intermittent videos. To evaluate its efficacy, we conducted a comparative analysis of the proposed model against two established algorithms: the DCP algorithm and the BI-GAN algorithm. These algorithms were tested on two distinct datasets: intermittent video sequences and the SBIdataset. Our analysis reveals that the proposed model significantly outperforms both the DCP and BI-GAN algorithms. Specifically, it achieves a performance improvement of 15:1 compared to the DCP algorithm and 28:2 compared to the BI-GAN algorithm. While the proposed algorithm may not represent the most technologically advanced method, its performance gains underscore its value and significance within the academic research community. By offering a robust solution to the complex challenge of intermittent video background reconstruction, this work contributes a valuable tool for researchers in the field.