Safaeid Hossain Arib
Machine Learning Engineer & Researcher
|
|
Hello, I am Safaeid Hossain Arib, a Machine Learning Engineer at ACI PLC. My work focuses on translating advances in machine
learning into dependable systems for real-world applications. Previously, I worked as an AI Engineer
at Ontik Technology, developing agentic AI and applied
machine learning solutions.
My current research investigates privileged-query attention distillation for teaching language models
where to attend. I conduct this work with guidance from Dr. Md. Mofijul Islam (AWS GenAI) and Dr. Tasnim Mohiuddin (Qatar Computing
Research Institute).
I hold a B.Sc. in Robotics and Mechatronics Engineering from the University of Dhaka. After completing my degree, I worked with Dr. Md Mehedi Hasan at the AVIS Research Lab on an open-world amodal counting
framework. During my undergraduate studies, I worked with Dr. Shafin Rahman and Dr. Sejuti Rahman on
continuous sign language translation.
When I’m not coding or reading research papers, you’ll likely find me on the basketball court, lifting
at the gym, or off exploring new places. I love connecting with curious minds—feel free to reach out!
I am actively seeking MS/PhD opportunities for Fall 2027.
Research interests
Reinforcement Learning
Multimodal Learning
Vision-Language Models
Embodied AI
Language Model Post-Training
Computer Vision
Connect with me
|
|
News and Updates
April 30, 2026:
Our paper "LAGEA: Language Guided Embodied Agents for Robotic
Manipulation" has been accepted at ICML
2026.
September 01, 2025:
Excited to begin my journey as a Machine Learning
Engineer at ACI
PLC.
February 14, 2025:
Our paper "SignFormer-GCN" is officially
published in PLOS ONE.
January 24, 2024:
Successfully defended my Undergraduate Thesis — a huge
milestone in my academic journey.
January 28, 2024:
Kicked off my career in the tech industry by joining Ontik
Technology as an AI Engineer.
September 25, 2021:
Honored to virtually present my poster "Classical
machine
learning approach for human activity recognition using location data" at Proc. UbiComp/ISWC
2021.
|
Research
My research spans reinforcement learning, multimodal learning, and post-training for vision-language
and embodied models, with a focus on grounded reasoning and adaptation.
|
Selected Publications and Preprints
|
|
CountOCC: Counting Through Occlusion for Open-World Amodal Counting
Safaeid Hossain Arib, Rabeya Akter, Abdul Monaf Chowdhury, Md Jubair Ahmed Sourov, Md Mehedi Hasan
Submitted to WACV 2027
[arxiv]
[code]
CountOCC reconstructs occluded object features through hierarchical multimodal guidance and attention-consistency distillation. It delivers state-of-the-art amodal counting under structured and unstructured occlusion and generalizes strongly across FSC-147, CARPK, and CAPTURE Real.
|
|
LAGEA: Language Guided Embodied Agents for Robotic Manipulation
Abdul Monaf Chowdhury, Akm Moshiur Rahman Mazumder, Safaeid Hossain Arib, Rabeya Akter
Accepted at ICML 2026
[arxiv]
[code]
LAGEA converts structured vision-language model reflections into temporally grounded reward signals for robotic manipulation. By localizing decisive moments and aligning language feedback with visual states, it improves success rates and convergence on the Meta-World MT10 benchmark.
|
|
T3Time: Tri-Modal Time Series Forecasting via Adaptive Multi-Head Alignment and Residual Fusion
Abdul Monaf Chowdhury, Rabeya Akter, Safaeid Hossain Arib
Accepted at AAAI 2026
[arxiv]
[code]
T3Time jointly models temporal, spectral, and semantic information for multivariate time-series forecasting. Horizon-aware gating and adaptive cross-modal alignment improve accuracy across eight benchmarks while retaining strong generalization in few-shot settings.
|
|
SignFormer-GCN : Continuous Sign Language Translation Using Spatio-Temporal Graph Convolutional Networks
Safaeid Hossain Arib, Rabeya Akter, Sejuti Rahman, Shafin Rahman
PLOS ONE
[paper]
[code]
SignFormer-GCN combines RGB video features with skeletal keypoints through a transformer and a spatio-temporal graph convolutional network. The joint representation captures complementary appearance and motion cues and achieves competitive translation performance across Bangla, American, and German sign language benchmarks.
|
|
Classical machine learning approach for human activity recognition using location data
Safaeid Hossain Arib, Rabeya Akter, Omar Shahid, Md Atiqur Rahman Ahad
UbiComp/ISWC '21 Adjunct: Adjunct Proceedings of the 2021 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2021 ACM International Symposium on Wearable Computers
[paper]
[code]
This work uses statistical features and a random forest classifier to recognize eight locomotion and transportation activities from smartphone location data. The approach achieved 78.14% validation accuracy and a 78.28% weighted F1 score in the SHL Challenge 2021 setting.
|
|
Bornil : An open-source sign language data crowdsourcing platform for AI enabled dialect-agnostic communication
Shahriar Elahi Dhruvo, Mohammad Akhlaqur Rahman, Manash Kumar Mandal, Md Istiak Hossain Shihab, A. A. Ansary, Kaneez Fatema Shithi, Sanjida Khanom, Rabeya Akter, Safaeid Hossain Arib, M.N. Ansary, Sazia Mehnaz, Rezwana Sultana, Sejuti Rahman, Sayma Sultana Chowdhury, Sabbir Ahmed Chowdhury, Farig Sadeque, Asif Sushmit
[arxiv]
Bornil is an open-source platform for crowdsourcing, annotating, and validating multilingual sign language data. The platform supported the creation of BornilDB v1.0, a large Bangladeshi Sign Language dataset, and established recognition benchmarks for this low-resource setting.
|
Teaching Language Models Where to Look with Privileged-Query Attention Distillation
Safaeid Hossain Arib, Rabeya Akter, Ishmam Nur Swapnil, Md Faiyaz Abdullah Sayeedi, Tasnim Mohiuddin, Md Mofijul Islam
Manuscript in preparation
This work studies how privileged queries can transfer attention supervision to language models, helping them identify the most relevant evidence during post-training without requiring privileged information at inference time.
|
Projects
Selected projects completed through professional work and academic coursework.
|
|
Pivotal Conversations
Built a business process automation system that analyzes historical social media data for content creators, generating monthly insight-driven reports with personalized recommendations and visualizations based on statistical analysis. This agentic workflow significantly boosted client viewership, streamlined reporting, eliminated manual effort, and enhanced decision-making.
Tools Used: pytorch, supermetrics, airtable, crewai
|
|
LookupMed
Developed a tool enabling patients to scan prescriptions, schedule medication dosages, and find alternative medicines. Leveraged an OCR and post‑processing pipeline to extract and validate drug names from prescription scans—ensuring alignment with reference data and achieving high accuracy. Collaborated with Dr. Javed Mostafa, Dean of the Faculty of Information at the University of Toronto, to pilot the system and validate its clinical and technical workflows.
|
|
LEGACY UNLOCKED
Designed and implemented an agentic workflow to extract and validate data from binary COBOL files into structured formats (CSV/TSV), ensuring consistency with copybook definitions. Successfully delivered demos to major consultancy firms like Oliver Wyman, showcasing how AI and agentic workflows can modernize legacy systems by unlocking data, enabling migration, generating insights, and supporting natural language interaction with enterprise data.
Tools Used: python, pandas, numpy, crewai
|
|
Automated Stock Trading Using Approximate Q Learning
RME 3211 (Intelligent Systems and Robotics Lab)
[code]
Developed an automated stock trading system using approximate Q-learning to optimize buy, sell, and hold
decisions for three Bangladeshi stocks, achieving an 11% return on investment and demonstrating the effectiveness
of reinforcement learning in financial decision-making.
|
|
A Comprehensive Comparative Analysis of Emotional Support Delivery by NAO Robots and Humans Across Varied Emotional States
RME 4211 (Human Robot Interaction Lab)
[code]
Conducted a comparative analysis of emotional support delivery by NAO robot and humans across 3 emotional
states, developing a system and questionnaire to quantify emotional impact, comfort levels, communication clarity,
and overall satisfaction, with the goal of assessing the efficacy of robotic versus human emotional support and
identifying areas for improvement.
|
Curriculum Vitae
For a detailed overview of my research, publications, experience, and service, please view my
current CV.
View CV (PDF)
|
Education
Below is a concise summary of my academic background to date.
|
|
■ Bachelor of Science in Robotics & Mechatronics
Engineering [2019–2024]
Department of Robotics & Mechatronics Engineering,
University of Dhaka,
Dhaka, Bangladesh
- CGPA: 3.66/4.00 (Final Year: 3.77/4.00)
-Final Year Project: Continuous Bangla Sign Language Translation: Mitigating the
Expense of Gloss Annotation with the Assistance of Graph [Presentation]
- Selected Courses: Artificial Intelligence, Machine Learning, Digital Image
Processing and Robot Vision, Intelligent Systems and Robotics, Human-Robot Interaction, Linear
Algebra, Multivariate and Vector Calculus, and Statistics for Engineers
■ Higher Secondary School Certificate (HSC) [2016–2018]
Notre Dame College,
Dhaka, Bangladesh
- GPA: 5.00/5.00
■ Secondary School Certificate (SSC) [2008–2016]
Dhaka Residential Model College,
Dhaka, Bangladesh
- GPA: 5.00/5.00
|
Experience
Selected research, industry, and teaching experience.
|
Research Experience
Research Assistant
Jun 2025 – Present
Department of Robotics and Mechatronics Engineering, University of Dhaka
Supervisors: Dr. Md Mehedi Hasan and Md. Jubair Ahmed Sourov
- Designed and implemented CountOCC, an open-world amodal counting framework, and developed
occlusion-augmented evaluation datasets for structured and unstructured scenes.
- Co-authored the resulting manuscript, submitted to WACV 2027.
Undergraduate Research Assistant
Jan 2023 – Dec 2023
Department of Robotics and Mechatronics Engineering, University of Dhaka
Supervisors: Dr. Shafin Rahman and Dr. Sejuti Rahman
- Designed and implemented a novel architecture for continuous sign language translation across
Bangla, American, and German benchmarks.
- Co-authored the resulting work, published in PLOS ONE.
|
Industry Experience
Machine Learning Engineer
Sep 2025 – Present
- Developed a few-shot classification pipeline for fine-grained recognition of visually similar
products.
- Building a lightweight text-to-speech model for few-shot voice cloning and low-resource language
adaptation.
- Improving a video-based truck identification pipeline by enhancing low-resolution license plate
imagery for more reliable detection and OCR.
- Developing an agentic assistant for meeting scheduling, document summarization, and conversational
search across company files through Drive and Calendar integration.
AI Engineer
Feb 2024 – Aug 2025
- Developed agentic B2B automation systems that generated social-media insights and automated
client-ready reporting, substantially reducing manual workflows.
- Built agentic ETL pipelines for COBOL copybook parsing and legacy modernization, enabling
AI-driven analytics and supporting demonstrations to consulting firms including Oliver Wyman.
- Developed a patient-facing OCR tool for prescription scanning, medication scheduling, and
alternative-medicine lookup under Dr. Javed Mostafa.
|
Teaching Experience
Instructor
Jun 2024 – Aug 2024
National Camp, Bangladesh Artificial Intelligence Olympiad
- Designed and delivered lessons on classification algorithms and evaluation metrics for high
school students and created quizzes to assess their understanding.
|
Awards and Scholarships
Over the years, I’ve been honored with a few recognitions—made possible by the invaluable support of
mentors and peers who have greatly influenced my growth.
|
■ Research Grant
- Special Innovation Fund, Bangladesh Ministry of Science and Technology (MoST),
Fiscal Year 2023–2024
Project ID: SRG-232431
■ Hackathons
- Honourable Mention, NASA Earth Observatory (EO) Dashboard Hackathon 2021
Organized by NASA, ESA, and JAXA
- 11th Place, Sussex-Huawei Locomotion Challenge 2021
Organized by University of Sussex and Huawei
|
Leadership
I dedicate part of my time to serving communities. Here are a few
selected contributions.
|
|
Student Activity Secretary
Jan 2023 – Dec 2023
IEEE Robotics and Automation Society Student Branch, University of Dhaka
- Organized workshops, interactive sessions, and experience-sharing sessions with industry
experts.
Mentor
Jul 2019 – Dec 2020
Bangladesh Robot Olympiad (BdRO)
- Conducted robotics workshops in multiple high schools and developed educational quizzes.
|
|
