EDN Dataset Analysis and Methodological Comparison: A Reproducibility Study

The EDN dataset comprises 5 subjects, with training and testing image dimensions set at 1024ￗ512. However, our method adopts a similar approach to iPER and ComplexMotion, utilizing image sizes of 512ￗ512 for both training and testing. The study conducted by Ghafoor et al. (2022) \cite{ghafoor2022walk} holds significant importance and demonstrates impressive results. Despite our inability to locate the code, we will endeavor to reproduce it and conduct comparative analyses. Furthermore, we will incorporate the recommended references to enhance the comprehensiveness of our research.