The videos captured by camera from laparoscopic surgery are prone to various distortions and have impact on visual quality. A state-of-the-art deep learning model called vision transformer was used to extract informative features by transferring learning and representation from the domain of natural images to the domain of laparoscopic videos. The proposed solution can help surgeons to avoid errors during laparoscopic surgery and can reduce time required for troubleshooting to do changes to the technical equipment that cause distortions.

An Artificial Intelligence (AI)-powered document extraction service can accelerate business processes by automating information extraction. The solution proposed AI-powered software that can read documents (pdf or picture) such as invoice or bills and convert them to useful data for decision making. The solution can simultaneously extract layout and structures of tables via a form recognition engine, and extract digital text via an OCR engine.