Segregation of Solid Waste Using Artificial Intelligence
rushikesh jachak (~rushikesh) |
Description:
The number of landfills has grown substantially and waste is being dumped at landfill without proper segregation. Manual segregation process is inefficient and unable to cope up with heavy amount of waste generated by growth in population. Examining and detecting the type of garbage using Image Classification can be extremely productive and less time consuming.
And hence, we developed a framework which automates the process of Segregating Waste into Biodegradable and Non-Biodegradable using Image Processing and Artificial Intelligence. This can help to improvise recycling and reusing of waste and henceforth increasing production of Manure and Renewable Energy through Solid Waste.
*The Outline of Tutorial is as Follows :*
Problem of Solid Waste Management - 5 minutes
Introduction and Some Statistics Related to Solid Waste in India along with the problems.
Related Work 5 Minutes
Earlier Research Work proposed to segregate solid Waste .
- How Deep Learning Can Effectively Segregate Waste - 5 Minutes
How Deep Learning is able to segregate this generated solid waste in more efficient manner.
- What is Region Convolutional Neural Network aka R-CNN. -5 Minutes
Introduction to R-CNN(R-CNN,Fast R-CNN,Faster R-CNN) and Working of IT.
- Our Framework For Segregation of Solid Waste Using Region - Convolutional Neural Network - 10 Minutes
We have created our own framework using Faster R-CNN to Classify waste into Biodegradable and Non-Biodegradable.
- Results, Future Work and Queries - 10 Minutes
Prerequisites:
Knowledge of Image Processing is good to know, but as such there are no prerequisites as the idea is more about application of Deep Learning Models in Real World.
Speaker Info:
We have already Presented this Work at ICACCP, 2019 under Computer Vision Track which was sponsored by Department of Science and Technology, India and SERB, India. The Research Work is soon going to be published in IEEE Xplore by end of September.
We have a past Experience of Speakers at Scipy International Conference, FOSSEE 2018 at IIT-Bombay where we conducted a workshop on Apache Spark and Its Implementation on Real World Datasets. Along with it, we also delivered a talk on PyDICOM - A Medical Imaging Library in Python.
We have Experience of more than 5 Hackathons, where our team has been into finale for all of the 5. Winners of Smart India Hackathon 2019 and AIR 6 at Smart India Hackathon 2018. We have been awarded with Most Innovative Solution Award by Deloitte at SIH 2018 and IOT based Smart Waste Management System by Cognizant at SIH 2019.
Currently Incubated under Ankur For Solving Problem of Solid Waste at Aurangabad.