NTU Wins 3rd Place in 2018 IEEE Conference on Image Processing

NTU added another feather to its cap. In the historic city of Athens, Greece, NTU_MiRA, a team made up of undergraduate and graduate students majoring in electrical engineering and computer science, was awarded 2nd Runner Up in the 2018 IEEE Signal Processing Society Video and Image Processing (VIP) Cup.

Medical imaging data has become a prominent research focus in medicine, electrical engineering, and computer science. This new focus can be attributed to recent advancements in machine learning technology, particularly in the area of deep learning, which enables significant breakthroughs in the analysis and generation of images, speech, and words. NTU’s College of Electrical Engineering and Computer Science (EECS), one of the innovation leaders in the group, continues to deliver and make waves in transdisciplinary studies in medicine and EECS.

The team from NTU’s College of EECS was awarded 2nd Runner Up in the 2018 VIP Cup, a competition hosted by the IEEE Signal Processing Society, for delivering identification accuracy and technological innovation for this year’s topic, “Lung Cancer Radiomics - Tumor Region Segmentation.” The team traveled to Athens, Greece for the presentation and commendation ceremony.

NTU_MiRA was beset by numerous challenges from the very beginning, but the students were determined and made stepping stones out of these frustrations. The most daunting part of the project was that such tumors could appear in any size, any form, and any place in the pulmonary region. This year’s competition also dictated that only limited training data would be allowed for the actual research, and this restriction made the team undertake numerous trials.

Led by Prof. Winston Hsu (徐宏民) of the Department of Computer Science and Information Engineering, the team was officially formed in June, and soon courageously ventured into new frontiers in the design of convolutional neural networks (particularly deep learning models for image recognition). The team also applied its research findings to medical science, hoping to identify and develop innovative solutions for the challenge.

Prof. Hsu also offers his mentoring skills to the Communications & Multimedia Laboratory (CMLab) at NTU, a project committed to the innovation of human-centric media technology. CMLab collaborates extensively with renowned partners, such as Nvidia and Microsoft Research Asia (MSRA). It is also an active member in projects spearheaded by the “AI Innovation Research Center” programs under the Ministry of Science and Technology.

This article is featured in No. 69 of NTU Highlights (December, 2018).