Faculty Seminar: Dr. Shaikh A. Fattah

Faculty Seminar: Dr. Shaikh A. Fattah

facultyseminar faculty seminar

participants

Speaker: Dr. Shaikh A. Fattah, Professor, EEE, BUET

Title of the Talk: Multi-perspective Deep Learning Architectures for Computer Vision and Biomedical Applications

Date: 12 May 2025

Time: 3 PM

Venue: Room number 634, ECE building, West Palashi, BUET, Dhaka-1000 The dept. of EEE, BUET is conducting regular research seminar for fostering knowledge sharing environment and encouraging collaborative research effort. This will be a platform for EEE, BUET faculty member, members of academia from home and abroad as well as experts in the field of electrical engineering to share their experience in research and innovation.

Abstract of the talk: Machine learning (ML) and deep learning (DL) models are extensively used for processing sensor data in tasks like disease detection, surveillance, human-machine interaction, and autonomous driving. A key challenge in designing these models is capturing the underlying physical characteristics of 1D/2D signals. In our attention-based, multi-perspective deep neural network architectures, we integrated temporal, channel, spatial, and guided attention mechanisms to enhance feature discrimination across various signal types, including PCG (heart sound), gyroscope, accelerometer, EEG, cardiac signals, and medical images, such as CT, MRI, ultrasound and endoscopy. In computer vision, depth estimation from monocular vision data is a challenging task where we achieve better performance with our proposed dual window transformer-based network that explores both local and global features. In many real life applications, users prefer to consult multiple sources to get better prediction. Effect of use of multi-expert learning for obtaining more explainable prediction will be addressed in this talk. Quantum machine learning (QML) is getting much attention day by days because of the exceptional computational abilities of quantum computers. Some of our recently developed QML architectures will be discussed which offer better expressibility and entangling capability for classifying speech and image data.

Bio of the speaker:  Dr. Shaikh Fattah received Ph.D. degree in ECE from Concordia University, Canada and later he was a visiting Postdoc at Princeton University, NJ, USA. He received B.Sc. and M.Sc. degrees from BUET, Bangladesh, where he is a Professor, Department of EEE. He published 285 international journal/conference papers and delivered 100+ Keynote/invited talks. His major research areas are machine/deep learning, robotics, computer vision, signal processing, and biomedical engineering. He served as Chair/Vice-Chair of IEEE RAS, EMBS, SSIT, IAS and PES Bangladesh chapters. He was IEEE Bangladesh Section Chair (2015-2016) and Chair of IEEE Region-10 AdHoc on Climate Change Sustainability (2023-24). He was the founding Chair of IEEE PES humanitarian activity committee (HAC) and Chair of IEEE SSIT Chapters. His student teams from BUET secured top positions in IEEE SP-Cup and VIP-Cup. He served IEEE Smart Village, HAC, TAB/PSPB PPSC, EAB, and SIGHT. He served as General Chair and TPC Chair of many IEEE Conferences, e.g. TENSYMP-2020, 27th ICCIT-2024. He received several awards, e.g. Concordia University’s Distinguished Doctoral Dissertation Prize (ENS, 2009), 2007 URSI Canadian Young Scientist Award, Dr. Rashid Gold Medal (M.Sc.), BAS-TWAS Young Scientists Prize (2014), 2016 MGA Achievement Award, 2017 Region-10 HTA Outstanding Volunteer Award and 2018 Region-10 Outstanding Volunteer Award. He was Editor-in-Chief of IEEE PES Enews and editorial board member of IEEE Potentials. He is Senior Editor of IEEE Access, Senior Member of IEEE and a Fellow of IEB.

Previous Post

Back to Top ↑