My Recent Research On Multimedia Processing

KSP Season 13 Episode 3:
My Recent Research On Multimedia Processing

Writer: Winda A.U.

Photographer: Winda A.U.

This KSP episode has the theme of multimedia processing and is presented by Professor Chuan-kai Yang from National Taiwan University. Multimedia processing that will be explained is divided into 4 categories, including image processing, video processing, graphics and visualization and applications for taking pictures.

Professor Chuan-kai Yang released 3 journals on Image Processing. The first journal is Mobile System Using AR To Display User's Password which discusses how to get a password to log into a website through facial recognition. The second journal is Object-Locating via Object Recognition and Indoor Positioning which discusses how to get the relative position and orientation of an object. The third journal is Privacy Protection and Beautification of Cornea Images which discusses how to protect information and privacy from the reflection of shadows on someone's cornea through photos of that person.

Several researches on Video Processing conducted by Professor Chuan-kai Yang include Video Length Reduction which functions to reduce video duration through surface carving, Video Retrieval By Trajectory or Appearance which functions to obtain the direction of motion of an object in the video, Encrypted-Domain Image /Video Retrieval which functions to secure the privacy of image and video data in the cloud through encryption methods, Transition Motion Synthesis for Video-Based Text To ASL (American Sign Language) which functions for translation of American sign language, and Automatic Generation of Video Navigation for Google Street View Data With Car Detection and Inpainting which serves to generate navigation videos using images from Google Street View.

There is some research conducted by Professor Chuan-kai Yang on the topic of Graphics and Visualization. Some of these researches include Interior Design to provide recommendations for furniture that can be placed in certain locations when doing interior design. Motion Learning that can generate motion from existing real examples, Paper Visualization, Matrix Visualization and Encryption-domain volume rendering.

Next is the application that will be used for rating photography competition scores. The factor used to judge an image is the composition technique to determine whether a photo meets the standards of a photo that has a good composition. The second application is “A Multi-person Selfie System Via Augmented Reality”. For example, Budi and Sarah want to take a photo together, they can take photos alternately and then the photos will be processed in such a way that the photos appear as if they were taken at the same time.

Thanks for sharing Professor Chuan-kai Yang, multimedia processing is undeniably a very interesting science to learn. It is hoped that KSP participants can understand multimedia processing and can apply it to their research in the future.