Deep Learning and Convolution is widely used for diagnosing Autism Spectrum Disorder (ASD), which is insufficient and computationally heavy. So, we introduce a hybrid model combining convolution and involution approach in the neural network, which demonstrates its effectiveness in classifying eye-tracking data. Our proposed model outperforms previous methods on this problem being the smallest model in comparison.

Title: Involution Fused ConvNet for Classifying Eye-Tracking Patterns of Children with Autism Spectrum Disorder

A work from Md. Farhadul Islam, Meem Arafat Manab, Joyanta J. Mondal, Sarah Zabeen Fardin Bin Rahman, Zadid Hasan, Farig Sadeque, and Jannatun Noor.

Link: https://arxiv.org/abs/2401.03575

Our research work titled “Uncovering local aggregated air quality index with smartphone captured images leveraging efficient deep convolutional neural network" has been accepted for publication as a Journal Paper in Nature Scientific Reports which is a Q1 Journal and one of the best journals for Multidisciplinary works in the world, having an impact factor of 4.6.

Scientific Reports is the 5th most-cited journal in the world, with more than 738,000 citations in 2022*, and receives widespread attention in policy documents and the media. This journal also ranks 37th (all categories) in terms of h-index on Google Scholar having an h-index of 210.

Our student research paper titled "Fog-Resilient Bangla Car Plate Recognition using Dark Channel Prior and YOLO," is live on CVF Open Access. Click the link below to go our paper.

Link: WACV 2024 Open Access Repository (thecvf.com) 

Our student research paper titled "Fog-Resilient Bangla Car Plate Recognition using Dark Channel Prior and YOLO," has been accepted at WVLL (WACV 2024 Workshop on Vision-based Understanding for Low-resource Languages)! IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) is the premier international computer vision event, comprising the main conference and several co-located workshops and tutorials. WVLL is a workshop at WACV 2024 focusing on vision-based applications for languages with limited linguistic resources.

This is my first co-supervised paper that got accepted after getting the RA position at BRAC University! Thanks to my students and supervisor.

Research Topic: 

Sustainable Deep Learning Applications and Approaches. 

My research is based on resource efficiency of deep learning approaches and sustainable applications for social development.

Lab: 

Computing for Sustainability and Social Good (C2SG) 

Will be managing C2SG Lab and co-supervising students and undergraduate members of the lab.

Supervisor: 

Dr. Jannatun Noor, Assistant Professor, BRAC University.

The abstract of our paper titled "Adversarial Attack against A.I. for Pedagogical Purposes: Designing Prompts for Students and not ChatGPT " got accepted at First Annual Elkana Symposium on Reimagining Teaching and Learning (Vienna, Austria).

I have received a travel and accommodation grant for presenting this paper. Really excited to visit Vienna, Austria for the first time.

Amount: 1950 Euros. 

Link to the conference.

BRAC University and the National Institute of Nuclear Medicine & Allied Sciences (Ninmas) of Bangladesh Atomic Energy Commission organized a meeting on the latter’s premises at Bangabandhu Sheikh Mujib Medical University (BSMMU) on 7 February, 2023 to explore avenues of possible collaboration. The delegation from BRAC University comprised Meem Arafat Manab, Lecturer of the Department of Computer Science and Engineering under the School of Data and Sciences and Manager of OSUN Science Shop under the School of General Education at BRAC University, Farhadul Islam Fuad, Teaching Assistant, and Joyanta Jyoti Mondal, Research Assistant.

Click for more.