Indian Students Solve Real-world Problems in Pilot Project, Technergize

POSTED BY on 06.26.2018

Can students really solve the biggest issues facing humanity today? Can they accomplish this using computer technologies?

The IEEE Computer Society India Council Student Activities Committee believes the answer to both of these questions is “yes.”

In 2017 and 2018, Richard E. Merwin scholar Shivam initiated a project titled Technergize that put these questions to the test. Technergize is a cross between an internship and competition that provides participants the opportunity to work on a substantial issue, drawing on innovative technologies, such as artificial intelligence, machine learning, network security, cyber security, natural language processing and data mining.

Secretary of the IEEE Computer Society India Students Activities Committee Srishti coordinated the project, which held its final round in association with IEEE Computer Society India Symposium 2018 (IEEE CSIS 2018).  The projects students worked on included:

  • Breast Cancer Phase Detection Using Histopathological Images
  • Cooperative 3D Map Generation Using Multiple Robots
  • Healthcare Data Access and Permission Management Using Ethereum Blockchains
  • Advanced Driver Assistance System for Indian Driving Conditions
  • Identification of Application Level DDoS Attacks
  • Design and Development of Efficient SILARE: SIGN Language Recognizer Using Image Mining Techniques

Breast Cancer Phase Detection Using Histopathological Images

The students who attacked the problem of breast cancer phase detection worked to identify whether a given set of histopathological images was cancerous, and what phase of development the cancer was in. The main challenge for the students was dealing with the inherent complexity of histopathological images.

Success in the project would help the medical community address a disease that affects 1.5 million women each year.

Cooperative 3D Map Generation Using Multiple Robots

The project utilized the integration of multiple robots to generate a 3D map of an unknown environment. The student team fitted each robot with a stereo vision camera with an IMU (inertial measurement unit) wheel.

The project’s real world application includes remote surveying of areas, such as underground mines, tunnels, caves or channels. During the highly influential discovery of Homo naledi in the Rising Star Cave System (South Africa) in 2016, the team used 3D mapping technology. Robotic mapping would have enhanced their ability to quickly map and analyze the cave and fossils.

Healthcare Data Access and Permission Management Using Ethereum Blockchains

Students working on this project tackled the challenge of the decentralization of health records, and made the case that blockchain for health care has the potential to standardize secure data exchange in a less burdensome way than previous approaches.

The students noted that blockchain acquired medical information and casualties of people in a region and could be monitored every week. It assisted authorities in keeping track of the actual rate of the spread of diseases. They also explained that blockchain provided a secure method for people to donate money to those in need without the intervention of formal charity organizations.

Advanced Driver Assistance System for Indian Driving Conditions

Students designed a system that assisted drivers in particularly challenging driving conditions in India, using Deep Learning and Artificial Intelligence. The students focused on the issues of pedestrian detection, safety alerts, the development of indicators for driving ability, and recognition of a drowsy driver.

The pedestrian detection system used advanced technology and algorithms to track human movements on the road and alerted the driver to help him or her keep from endangering people on the road. The Driver Drowsiness System determined if a driver was drowsy while driving and awakened him or her with a shock. The Driver Quality Indicator determined driving ability. And the safety alert system defined a safe distance and safe speed based on the particular situation of the vehicle.

Identification of Application Level DDoS Attacks

Periodically, a story makes the news about a website’s being hit by distributed denial of service attack (DDoS) in which someone overwhelms a system in order to deny its users service.

A group of students sought a solution for this problem through their Technergize project. Since most methods of countering or protecting from DDoS suffer from either efficiency or accuracy shortcomings, the students introduced a conglomeration of Software Defined Networking (SDN) and deep learning techniques to overcome the issues. They incorporated a combinatorial approach for the detection process by merging deep learning technology and SDN backed by cryptographic schemes to ensure efficiency with security.

The students further discovered that there was a trade-off between detection speed and adaptability, and that they could tune the solution analytically to provide a pragmatic approach for the detection of DDoS attacks in an application layer and consequently pave the way for its mitigation.

Design and Development of Efficient SILARE: SIGN Language Recognizer Using Image Mining Techniques

The group of students worked with a series of neural networks to construct a framework for interpreting American Sign Language to aid communication between members of the deaf and the hearing communities in the United States.

The group designed their algorithms to train the recognition models with minimal human input, so they could improve on the results obtained by previous research and other sign-language recognition. They focused on using machine learning concepts to overcome computational accuracy.

Next Steps

The team conducting the project on Breast Cancer Phase Detection Using Histopathological Images led by Shrutina Agarwal came in first place this year, followed by the group led by Balan Sethuramalingam working on the issue of Design and Development of Efficient SILARE: SIGN Language Recognizer using Image Mining Techniques.

With the conclusion of a successful pilot, the IEEE Computer Society India Council Student Activities Committee plans to extend Technergize in a way that—besides providing the students with an opportunity to work on real-world projects—would also help to solve problems in the rural parts of India, and this time, with even better and more substantial support from industry. Commenting on the humanitarian element of the project, Shivam felt it best to quote the Bhagavad Gita: “Man is made by his belief. As he believes, so he is.”

Any person seeking to join the project or to conduct something similar can contact Shivam at shivam.1996.in@ieee.org and Srishti at itzsrishti.smile@ieee.org for advice or recommendations.

 

This article was written by Anamitra Sarma, Content Co-Lead IEEE Computer Society India Student Activities Committee, based on a report provided by Srishti, Secretary IEEE Computer Society India Student Activities Committee.

POSTED IN  |  TAGS:   •    •    •    •  

Leave a Reply

Your email address will not be published. Required fields are marked *