Minhajul Arifin Badhon

Computer Vision Research Developer
Wave9 Technologies Inc.
Regina, SK, Canada
September 2019 - Present

Hello! I graduated from the Department of Computer Science and Engineering, Military Institute of Science and Technology in 2016. During my final year, I started working for a New Jersey, USA, based startup, QI Analysis Inc., as part time software engineer and joined the same company as full time software engineer just after my graduation.

I love coding. During my undergrad life, I participated in more than 15 national and regional programming contests representing my university. I acted as Co-Founder of MIST Computer Club and was elected as President of MIST Computer Club in 2015.

I also took part in various application development competitions, project showcasing and international robotics competition such as The University Rover Challenge (URC) and had mentionable achievements.

University of Saskatchewan

Computer Science

Military Institute of Science and Technology

Computer Science and Engineering (CSE)

CGPA: 3.79/4.00 (Final year GPA 3.98)

Notre Dame College

Science

GPA: 5.00/5.00

SOS Hermann Gmeiner College

Science

GPA: 5.00/5.00

IELTS

Band Score 8.0

Reading: 9.0, Listening: 8.5, Writing: 7.0, Speaking: 7.0

Computer Vision Research Developer

Wave9 Technologies Inc.

Software Engineer

QI Analysis Inc.

President

MIST Computer Club

Founding committee member of the computer club. Organized courses on contest programming and programming languages. Conducted courses on data structures and algorithms.

Internship

Banglalink Digital Communications Ltd. , Bangladesh

Assistant Secretary and Senior Instructor

MIST Computer Club

Taken courses on data structure and algorithm.

Team Member

MIST URC Team - Mongol Barota

Worked as a member of the team Mongol Barota that participated in "University Rover Challenge 2014" arranged by The Mars Society at the Mars Desert Research Station (MDRS) in the remote, barren desert of southern Utah, USA in late May, 2014. It was the first team to participate from Bangladesh. The team scored 187 points in 5 different scientific tasks and achieved 12th position out of 31 registered teams.

Co Founder and Junior Instructor

MIST Computer Club

Taken courses on C, C++ .

Committee Member

Google Developers Group Bangladesh

Helped Google Developers Group to organize Google DEVFEST Dhaka 2014.

Head Volunteer (Publicity)

MIST Computer Club

Designed first ever banner and poster for MIST Computer Club. Worked to raise awareness about programming and club activities among students.

Fast rotated bounding box annotations for object detection

Supervisor: Dr. Ian Stavness, Associate Professor, Department of CS, University of Saskatchewan

Traditionally, object detection models use a large amount of annotated data and axis-aligned bounding boxes (AABBs) are often chosen as the image annotation technique for both training and predictions. The purpose of annotating the objects in the images is to indicate the regions of interest with the corresponding labels. Accurate object annotations help the computer vision models to understand the distinct patterns of the image features to recognize and localize different classes of objects. However, AABBs are often a poor fit for elongated object instances. It’s also challenging to localize objects with AABBs in densely packed aerial images because of overlapping adjacent bounding boxes. Alternatively, using rectangular annotations that can be oriented diagonally, also known as rotated bounding boxes (RBB), can provide a much tighter fit for elongated objects and reduce the potential bounding box overlap between adjacent objects. However, RBBs are much more time-consuming and tedious to annotate than AABBs for large datasets. In this work, we propose a novel annotation tool named as FastRoLabelImg (Fast Rotated LabelImg) for producing high-quality RBB annotations with low time and effort. The tool generates accurate RBB proposals for objects of interest as the annotator makes progress through the dataset. It can also adapt available AABBs to generate RBB proposals. Furthermore, a multipoint box drawing system is provided to reduce manual RBB annotation time compared to the existing methods. Across three diverse datasets, we show that the proposal generation methods can achieve a maximum of 88.9% manual workload reduction. We also show that our proposed manual annotation method is twice as fast as the existing system with the same accuracy by conducting a participant study. Lastly, we publish the RBB annotations for two public datasets in order to motivate future research that will contribute in developing more competent object detection algorithms capable of RBB predictions.

Analysis of Big Data to study the pattern of students’ activities

Supervisor: Fahim Hassan Khan, Assistant Professor, Department of CSE, MIST

We analyzed the correlations of activity patterns among the students of different departments using unsupervised learning and developed a predictive model to give suggestion to students about which club will be more appropriate for them based on their interests and participations in different activities using supervised learning.

Automatic Microplot Localization Using UAV Images and a Hierarchical Image-Based Optimization Method

Plant Phenomics

To develop new crop varieties and monitor plant growth, health, and traits, automated analysis of aerial crop images is an attractive alternative to time-consuming manual inspection. To perform per-microplot phenotypic analysis, localizing and detecting individual microplots in an orthomosaic image of a field are major steps. Our algorithm uses an automatic initialization of the known field layout over the orthomosaic images in roughly the right position. Since the orthomosaic images are stitched from a large number of smaller images, there can be distortion causing microplot rows not to be entirely straight and the automatic initialization to not correctly position every microplot. To overcome this, we have developed a three-level hierarchical optimization method. First, the initial bounding box position is optimized using an objective function that maximizes the level of vegetation inside the area. Then, columns of microplots are repositioned, constrained by their expected spacing. Finally, the position of microplots is adjusted individually using an objective function that simultaneously maximizes the area of the microplot overlapping vegetation, minimizes spacing variance between microplots, and maximizes each microplot’s alignment relative to other microplots in the same row and column. The orthomosaics used in this study were obtained from multiple dates of canola and wheat breeding trials. The algorithm was able to detect 99.7% of microplots for canola and 99% for wheat. The automatically segmented microplots were compared to ground truth segmentations, resulting in an average DSC of 91.2% and 89.6% across all microplots and orthomosaics in the canola and wheat datasets.

Global Wheat Head Detection 2021: An Improved Dataset for Benchmarking Wheat Head Detection Methods

Plant Phenomics

The Global Wheat Head Detection (GWHD) dataset was created in 2020 and has assembled 193,634 labelled wheat heads from 4700 RGB images acquired from various acquisition platforms and 7 countries/institutions. With an associated competition hosted in Kaggle, GWHD_2020 has successfully attracted attention from both the computer vision and agricultural science communities. From this first experience, a few avenues for improvements have been identified regarding data size, head diversity, and label reliability. To address these issues, the 2020 dataset has been reexamined, relabeled, and complemented by adding 1722 images from 5 additional countries, allowing for 81,553 additional wheat heads. We now release in 2021 a new version of the Global Wheat Head Detection dataset, which is bigger, more diverse, and less noisy than the GWHD_2020 version.

Global Wheat Head Detection (GWHD) Dataset: A Large and Diverse Dataset of High-Resolution RGB-Labelled Images to Develop and Benchmark Wheat Head Detection Methods

Plant Phenomics

The detection of wheat heads in plant images is an important task for estimating pertinent wheat traits including head population density and head characteristics such as health, size, maturity stage, and the presence of awns. Several studies have developed methods for wheat head detection from high-resolution RGB imagery based on machine learning algorithms. However, these methods have generally been calibrated and validated on limited datasets. High variability in observational conditions, genotypic differences, development stages, and head orientation makes wheat head detection a challenge for computer vision. Further, possible blurring due to motion or wind and overlap between heads for dense populations make this task even more complex. Through a joint international collaborative effort, we have built a large, diverse, and well-labelled dataset of wheat images, called the Global Wheat Head Detection (GWHD) dataset. It contains 4700 high-resolution RGB images and 190000 labelled wheat heads collected from several countries around the world at different growth stages with a wide range of genotypes. Guidelines for image acquisition, associating minimum metadata to respect FAIR principles, and consistent head labelling methods are proposed when developing new head detection datasets. The GWHD dataset is publicly available at http://www.global-wheat.com/and aimed at developing and benchmarking methods for wheat head detection.

Light following and obstacle avoiding robot using autonomous and android based manual controller

2015 18th International Conference on Computer and Information Technology (ICCIT)

In this work, we have presented a robot, which is compact, autonomous and fully functional. It is a proposed model which can be used in such an environment, which may be vulnerable and risky to human being. It has three types of functions. Those are light following, obstacle detection and controlling from an android device through bluetooth module. A mobile robot having a control unit integrating the processing and the main sensor functionalities into an android device is described and demonstrated in this paper.

Mongol Barota - A Next Generation Rover

8th International Conference on Software, Knowledge, Information, Management and Application (SKIMA)

This paper scrutinizes Mongol Barota - a fully functional, stand-alone mobile platform rover which is capable to act as a human assistant to perform various scientific tasks in extreme adversities. The control system of the rover is designed in such a way that it can be commanded from a blind station within 1 kilometer range. It has successfully taken part in 8th annual University Rover Challenge organized by the Mars Society at the Mars Desert Research Station (MDRS) in the remote, barren desert of southern Utah, USA in late May, 2014. It has been traced out as the first entrance in this competition from Bangladesh and occupied 12th position out of 31 registered teams from 6 countries of 4 continents. The rover architecture maps the associated components to make it capable to perform the assigned tasks namely - Sample Return Task, Astronaut Assistance Task, Equipment Servicing Task and Terrain Traversing Task. Among these, the first task refers to search for the evidence to identify the existence of life after detailed analysis of collected soil sample from a selected site. In Equipment servicing task, rover has to perform a sequence of operations that mainly includes switching on a compressor and working with a series of pipes, hoses, valves and other such equipment. Astronaut assistance task intends the rover to collect tools from some given GPS locations and then delivery of each of them to the corresponding locations with provided GPS coordinates. Rover has to traverse an adverse terrain in order to pass through a set of target gates for completion of the terrain traversing task. This paper provides a detailed demonstration of the Mongol Barota rover, ins and outs of its architecture, facts and features, system components, logic, logistics and techniques adopted to implement several tasks representing its overall capabilities.

Hierarchical Image-based Optimization Method for Automatic Microplot Localization Using UAV-Acquired Images

5th Annual Plant Phenotyping and Imaging Research Center Symposium

Automatic microplot localization from uav-acquired images using an optimization method

Phenome, February 24–27, Tucson, AZ 2020

Towards (Real Time) automatic localization and labeling of field plots from drone imagery

Black in AI workshop, NeurIPS 2019

Pheno-Canola plot localization using an optimization method

4th Annual Plant Phenotyping and Imaging Research Center Symposium

Predicting the Impact of Field Topography on Yield of Canola: A Comparison of Linear Modelling and Machine Learning Approaches

4th Annual Plant Phenotyping and Imaging Research Center Symposium

Best Poster Award (Hierarchical Image-based Optimization Method for Automatic Microplot Localization Using UAV-Acquired Images)

5th Annual Plant Phenotyping and Imaging Research Center Symposium

Best Poster Award (Pheno-Canola plot localization using an optimization method)

4th Annual Plant Phenotyping and Imaging Research Center Symposium

Dean's Scholarship

University of Saskatchewan

MIST Dean's List 2015

Military Institute of Science and Technology

MIST Commandant's List 2014

Military Institute of Science and Technology

MIST Dean's List 2013

Military Institute of Science and Technology

Dhaka Board Scholarship

Board of Intermediate and Secondary Education, Dhaka

For the result of HSC.

Junior Board Scholarship

Dhaka Education Board

Talent-pool, Merit Position 16th

Primary Board Scholarship

Dhaka Education Board

Talent-pool, Merit position 3rd

Certificate of Achievement - The ACM ICPC Dhaka Site 2015

ACM International Collegiate Programming Contest

Secured 19th position. Team MIST_Endless_Code.

Certificate of Achievement - The ACM ICPC Dhaka Site 2014

ACM International Collegiate Programming Contest

Secured 11th position. Team MIST_ENDLESS_CODE.

MIST Best Project Award 2015 (Department of CSE)

Military Institute of Science and Technology

For the project named 'MIMO - An Intelligent Robot'

Champion in Project Showcasing - Mobile Application Development Course 2015

Military Institute of Science and Technology

Built android application named 'Lostpedia'.

MIST Best Project 2014

Military Institute of Science and Technology

Project: Mongol Barota ( Building a next generation Mars Rover )
Mongol Barota is a project that builds a next generation mars rover to assist human in Mars. This project was built to compete in the international robotics competition - University Rover Challenge 2014 and achieved 12th position among the 31 teams all over the world. It was the first mars rover built in Bangladesh.

Champion in Project Showcasing of DUITS ERICSSON NATIONAL CAMPUS IT FEST 2014

Dhaka University IT Society

COMMANDANT'S COMMENDATION - THE MARS SOCIETY'S UNIVERSITY ROVER CHALLENGE (URC) 2014, USA

Military Institute of Science and Technology

Commandant's commendation for securing 11th place in URC 2014.

University Rover Challenge 2014- Certificate of Participation

The Mars Society

Secured 11th place in URC 2014.