Abu Ibne Bayazid
"Improving Human-AI Interaction via Explainable Chatbot: In Case of COVID-19 Diagnosis"
We will make a chatbot where user can provide their clinical data,CT scan and or X-Ray images.
By providing these data, user will get the prediction made by the system whether they have COVID-19 or not. Moreover, they will get survival probability.
Most interesting part of this thesis is that user can ask different questions to the system.
For example, user can ask the explanation for the system's prediction result and many more.
Here, we will use single modality(each for clinical,CT and X-ray separately) and multimodality(clinical,CT and X-ray combinedly).
We will use random-forest, VGG-16,CNN,LIME,SHAP.
Technology: python for data analysis, FLASK for backend, VueJS for chatbot front end and dialogflow for chatbot training.
Link: Check out the master thesis proposal presentation
"Automated Walk-in Assistant for the Visually Impaired People"
From the early days of my secondary school level, I always thought how to solve the surrounding problems of us and how
to make the world better. I along with another member, worked to solve the problem of the physically impaired people to
walk more easily with the help of technology during my undergrad thesis work. We built a device which can detect real time
objects and obstacles not only in front of the user but also the obstacles and holes on the surface. It can alert the user about
obstacles and can guide the user about direction to move. Image processing was used for surface obstacles, holes detection and sensors
were used for front obstacle detections along with an android mobile device. This research project was accepted at electrical
engineering and information & communication technology (ICEEICT), 2014 ,International Conference and subsequently published in IEEE
Xplore. Another publication was accepted and presented at IEEE R10-HTC, 2017 and the title was Towards developing walking assistants
for the visually impaired people. This remarkable research work was published in almost all national newspapers of Bangladesh likewise
( http://www.thedailystar.net/through-the-eyes-of-technology-57229 ) and also telecasted in some TV channels where they mentioned about
this research project and how it will help the physically impaired people. Some of charity organizations who work for the blind people
also praised and contacted with our team. Our university’s vice chancellor himself appreciated this work. Furthermore, this work was
awarded by “Joy Bangla Award” in 2015 which is a government award from the ministry of ICT for the contribution in the field of
Publication1 at IEEE:
Publication2 at IEEE: