Wireless Capsule Endoscopy (WCE) is the preferred technology for the diagnosis and evaluation of small bowel disorders. The concluding WCE video comprises as many as 150,000 images. Within this extensive collection of images, the majority are classified as normal, with only a singular or a few frames identified as pathological. The transmission of such a substantial volume of data from the WCE is constrained by the device's limited battery capacity. The primary objective of this study is to minimize the computational load by shifting the maximum of computations at the decoder side while reducing the information transmitted from the encoder. Our project presents three primary contributions. First, a new Capsule Endoscopy framework leveraging Deep Learning (DL-CEndo) is proposed for the WCE compression system. The figure illustrates an overview of the proposed DL-CEndo framework. This framework aims to significantly decrease the transmission of unnecessary data by sending Low- Resolution (LR) luma images, thereby conserving energy in the WCE device. Second, a new colorization model named EndoColorGAN diffusion-based is implemented to reconstruct the colors of WCE images. To expedite the diagnostic review of the WCE video, we introduce a visualization demo on this page, specifically designed for fast and intelligent summarization, as defined by the Offline-Decoder in the figure below. At the physician’s workstation, and before the colorization process is done, the received video consists of Super Resolution Luma normal images and RGB pathological images. We have implemented an algorithm that enables physicians to visualize the pathological sequence, which constitutes the primary focus, before proceeding to colorize the normal images. This simplified process not only enhances efficiency but also provides critical information directly to the physician in a short time.