Team NTU – Tan Ying Hao, Shao Yiyang, Liu Siyuan, Wu Bian, Li Yuanrui (back), Ren Daxuan
A new Deep Learning challenge was added to the recent ISC 2017 Student Cluster Competition. As devised by Scot Schultz from the HPC Advisory Council, the “Solving the CAPTCHA Challenge and Achieving the Highest Degree of Model Accuracy” was won by Nanyang Technological University (NTU).
The ISC Student Cluster Competition provides an excellent opportunity for students to hone their HPC skills, for example building their own cluster and fine tuning HPC applications,” said Professor Francis Lee from NTU. The NTU team has brought honors to the university and Singapore as a co-winner of the Deep Learning Challenge, showing their innovativeness in the design and use of the DenseNet network. Participation in ISC SCC would not have been possible without the support and sponsorship from NSCC.”
According to Schultz, Captcha is an ideal kind of problem for Deep Learning:
“We decided to use a slightly more interesting use case of solving for Captcha because it not only highlights the power of deep learning to be a useful tool create models to recognize and classify unwieldy data, such as distorted characters, grainy images and overlapping characters, but it also demonstrates that it is possible for this powerful technology to be used in less positive ways, such as solving for security or privacy. Realizing that everyone has access to the tools we use to move society forward, we need to be aware of the possible mis-use, especially as it becomes more pervasive across industry, healthcare, financial services, and the like.”
In this video, Dan Olds from OrionX interviews Team NTU at ISC 2017.
This competition provided an informative and challenging opportunity for the team from NTU.
We not only learned from other teams, but were also able to share with them our knowledge and skill in areas such as HPCC (High Performance Computing Cluster), HPCG (High Performance Conjugate Gradients) and HPL (High Performance LINPACK)”, said Team Captain, Liu Siyuan.
