Collaboration in Research: Utilizing Deep Learning to Solve National Issues

Deep learning is the fastest-growing field in artificial intelligence, helping computers make sense of infinite amounts of data in the form of images, sounds, and text. Using multiple levels of neural networks, computers now have the capacity to see, learn, and react to complex situations as well or better than humans. This leads to a profoundly different way of thinking about big data, technology, products, and services.

There are many useful aspects of deep learning, such as in Internet & Cloud (image classification, speech recognition, language translation, language processing, sentiment analysis, and recommendation), medicine and biology (cancel cell detection, diabetic grading, and drug discovery), media & entertainment (video captioning, video search, and real-time translation), security & defense (face detection, video surveillance, and satellite imagery), and autonomous machines (pedestrian detection, lane tracking, and recognizing traffic signs). Those can be achieved because of the High-Performance Computing (HPC) and the deep neural network ability to process raw data into high-level features.

According to Dr. Bens Pardamean, Director of the Bioinformatics & Data Science Research Center (BDSRC), BINUS UNIVERSITY, it isn’t only a trend, but also a strategy to handle and analyze big data. “Big data has become a challenge not only for the commercial world, but also in research,” said Dr. Bens, during a Workshop and Seminar for Preparing Joint Research Proposals on Thursday, November 3rd, 2016.

To tackle that challenge, BDSRC collaborates with its partners from the Nationwide University Network in Indonesia (NUNI), especially universities with well-established computer science backgrounds. The collaboration aims to generate topics that have an impact on national issues.

There are many potential topics that can be related to the use of HPC in the domain – plantations, computer graphics/games, national border security, fishery, forest fires (hotspots), computer vision, big data analysis, etc.

Preparing for Joint Research Proposals

BINUS UNIVERSITY is committed to providing a world-class education with international research experience that makes positive contributions to the global community.  One important aspect is to collaborate with industries and other research institutions

BINUS UNIVERSITY with NVIDIA presented a Workshop and Seminar for Preparing Joint Research Proposals at BINUS FX Campus, Senayan, Jakarta, on Thursday, November 3rd, 2016. This event was designed to facilitate researchers of the Nationwide University Network in Indonesia (NUNI) to learn, establish a consortium, and prepare joint research proposals in applying deep learning methods in various domains (marine, security, plantation, agriculture, forestry, etc.) using High-Performance Computing (HPC).

There were 46 participants from 12 universities that attended this workshop and seminar. The participants came from BINUS UNIVERSITY, Unika Soegijapranata Semarang, Brawijaya University, Diponegoro University, Hasanuddin University, the University of Indonesia, Indonesia Islam University (UII), Atma Jaya Catholic University, Petra Christian University, Satya Wacana Christian University, Sanata Dharma University, Mataram University, and the Bogor Agriculture Institute (IPB).

The objectives of this workshop and seminar were to increase awareness and understanding of deep learning methods through applications in various domains (marine, security, plantation, agriculture, forestry, etc.) with High Performance Computing (HPC) to establish an HPC consortium and to prepare joint proposals.

Several sessions were held before the Workshop and Seminar for Preparing Joint Research Proposals and establishment of the HPC Consortium – a welcoming note & opening speech by Prof. Bahtiar Saleh Abbas, Ph.D., Vice Rector of Research and Technology Transfer, BINUS UNIVERSITY, a workshop & seminar overview by Dr. Bens Pardamean, GPU Technologies for Accelerated Computing (Introduction) by Ettikan Kandasamy Karuppiah, Ph.D., Director of Developers Ecosystem, APAC South NVIDIA, and Deep Learning & Machine Learning Using GPU: Use Cases by Ettikan Kandasamy Karuppiah.

From the workshop and seminar, nine topics were proposed: sleep monitoring, appropriate cropping patterns for specific regions, ancient manuscript translation, a genome-wide association study (GWAS) in human and agriculture studies, object recognition in marine, image compression, weather change-resistant plants, Indonesia’s tropical fruits classification, and protein interactions for diabetes mellitus.

Dr. Bens Pardamean expressed his hope that there will be some proposals that have an impact on national issues and the HOC consortium can be established to sustain this activity. “And maybe we need to establish a big data consortium too, so the community can also get involved,” he added.

Sustainability is the major challenge for this consortium, according to Dr. Bens. Therefore, the researchers have to be proactive to develop the proposals.

Active and consistent contributions from the researchers and members in developing proposals is crucial because proposals are the vital key to apply for research funding from funding organizations. “I am optimistic that this activity will be sustainable,” said Dr. Bens.

A Collaboration in Research with NVIDIA

NVIDIA is a world leader in visual computing and has been a pioneer in accelerating deep learning. NVIDIA has been developing deep learning software, libraries, and tools for a number of years. The NVIDIA invention, unique GPU-based platforms, are advancing a wide range of industries, from gaming, medicine, supercomputing, Cloud services, and autonomous driving. NVIDIA GeForce is synonymous with PC gaming and, with 200 million gamers, it’s the largest gaming platform in the world.

NVIDIA Tesla GPUs power the fastest supercomputers in the U.S., Europe, and Japan; and they were selected by the U.S. Dept. of Energy for two next-generation supercomputers. NVIDIA GPUs regularly enable breakthroughs in vital areas of research featured in leading scientific journals.

BINUS has been selected as an NVIDIA GPU Education Center from NVIDIA and has built a strong relationship with NVIDIA in the past three years. NVIDIA will also be one of the consortium’s main partners in utilizing deep learning.

“NVIDIA confirmed that this activity will be held annually. Not only will it be conducted at BINUS, it will also be held at other universities like Petra Christian University and Hasanuddin University,” Dr. Bens explained.

(ERON)