Mustansar Fiaz

Staff Research Scientist | Computer Vision | Deep Learning | AI
Mustansar Fiaz

Biography

I am a Staff Research Scientist at IBM Research, Abu Dhabi, UAE, where I work on advancing the frontiers of computer vision and deep learning. My research centers on visual–spatial and temporal perception, with an emphasis on explainable AI. My work focuses on developing innovative solutions spanning remote sensing, medical imaging, visual language models, and person search, emphasizing real-world deployment and practical impact.

Before joining IBM Research in October 2023, I was a Postdoctoral Researcher at Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), where I worked with Prof. Fahad Khan and Dr. Hisham Cholakkal at the Intelligent Visual Analytics Lab. During this time, I contributed to cutting-edge research in vision transformers, multi-modal learning, and efficient deep learning architectures.

I received my Ph.D. in Computer Science and Engineering from Kyungpook National University, Daegu, Republic of Korea in 2021 under the supervision of Prof. Soon Ki Jung, where my thesis on deep learning for visual tracking received the Outstanding Research CSE Thesis Award. Prior to that, I completed my Master's degree at Sejong University, Seoul, South Korea in 2016 under the guidance of Prof. Sung Wook Baik, and my Bachelor's degree in Computer and Information Science from Pakistan Institute of Engineering and Applied Sciences (PIEAS) in 2011.

News

Latest
July 2025
BMVC 2025

Paper accepted

Oral
March 2025
IEEE IGARSS 2025

Oral presentation

Top Venue
February 2025
CVPR 2025

Paper accepted

January 2025
IEEE ISBI 2025

Paper accepted

Oral
November 2024
WACV 2025

Oral presentation

July 2024
Pattern Recognition

Journal paper accepted

May 2024
US Patent Published

Person search technology

2 Papers
April 2024
IEEE IGARSS 2024

Both oral presentations

January 2024
ISBI 2024

Paper accepted

2 Papers
January 2024
IEEE TGRS

Two papers accepted

December 2023
IEEE TAI

Paper accepted

Career
October 2023
Joined IBM Research

Staff Research Scientist

Recent Publications

GeoVLM-R1: Reinforcement Fine-Tuning for Improved Remote Sensing Reasoning
Mustansar Fiaz*, H. Debary, P. Fraccaro, D. Paudel, L. V. Gool, F.S. Khan, S. Khan
Under Review
Paper Code
CLIP meets DINO for Tuning Zero-Shot Classifier using Unlabeled Image Collections
M. F. Imam, R. F. Marew, J. Hassan, Mustansar Fiaz, A. F. Aji, H. Cholakkal
BMVC 2025
Paper Code
HyRet-Change: A hybrid retentive network for remote sensing change detection
Mustansar Fiaz, M. Noman, H. Debary, K. Ali, H. Cholakkal
IEEE IGARSS 2025 (Oral)
Paper Code
EarthDial: Turning Multi-sensory Earth Observations to Interactive Dialogues
S. Soni*, A. Dudhane*, H. Debary*, Mustansar Fiaz*, M. A. Munir, M. S. Danisho, P. Fraccaro, C. Watson, L. J. Klein, S. Khan, F.S. Khan
CVPR 2025
Paper Code
COSNet: A Novel Semantic Segmentation Network using Enhanced Boundaries in Cluttered Scenes
M. Ali, M. Javaid, M. Noman, Mustansar Fiaz, S. Khan
WACV 2025 (Oral)
Paper Code
ChangeBind: A Hybrid Change Encoder for Remote Sensing Change Detection
Mustansar Fiaz, M. Noman, H. Cholakkal
IEEE IGARSS 2024 (Oral)
Paper Code
Guided-attention and gated-aggregation network for medical image segmentation
Mustansar Fiaz, M. Noman, H. Cholakkal, R.M. Anwer, J. Hanna, F.S. Khan
Pattern Recognition, 2024
Paper Code
ELGC-Net: Efficient Local–Global Context Aggregation for Remote Sensing Change Detection
M. Noman, Mustansar Fiaz, H. Cholakkal, S. Khan, F.S. Khan
IEEE TGRS, 2024
Paper Code
Remote Sensing Change Detection With Transformers Trained from Scratch
M. Noman, Mustansar Fiaz, H. Cholakkal, S. Narayan, R.M. Anwer, S. Khan, F.S. Khan
IEEE TGRS, 2024
Paper Code
SA2-Net: Scale-aware Attention Network for Microscopic Image Segmentation
Mustansar Fiaz, M. Heidari, R.M. Anwer, H. Cholakkal
BMVC 2023 (Oral)
Paper Code
SAT: Scale-Augmented Transformer for Person Search
Mustansar Fiaz, H. Cholakkal, R.M. Anwer, F. Khan
WACV 2023
Paper Code
PS-ARM: An End-to-End Attention-aware Relation Mixer Network for Person Search
Mustansar Fiaz, H. Cholakkal, S. Narayan, R.M. Anwer, F. Khan
ACCV 2022
Paper Code
View All Publications

Mentorship

  • Mubashir Noman (PhD @ MBZUAI, 2022-2025)
  • Muhammad Ali (PhD @ MBZUAI, 2022-2025)
  • Daniya Abdul Kareem (PhD @ MBZUAI, 2022-2025)
  • Mohamed Fazli Mohamed Imam (MS @ MBZUAI, 2022-2024)
  • Imane Hilal (MS @ MBZUAI, 2021-2023)
  • Mohammed Almansoori (MS @ MBZUAI, 2021-2023)
  • Reem Alameeri (MS @ MBZUAI, 2021-2023)
  • Md Maklachur Rahman (MS @ KNU, 2018-2020)

Service & Volunteers

  • ICCV 2023 Workshop on New Ideas in Vision Transformers, Paris, France, October 02, 2023
  • NeurIPS 2022 Workshop on Vision Transformers: Theory and Applications, New Orleans, USA, Dec 9, 2022
  • ACCV 2022 Workshop on Vision Transformers: Theory and Applications, Macau SAR, China, Dec. 5, 2022
  • IW-FCV 2021 The 27th International Workshop on Frontiers of Computer Vision, Daegu, South Korea, Feb. 22-23, 2021

Honors & Awards

  • Oral Paper Presentation in IEEE IGARSS 2025
  • Oral Paper Presentation in WACV 2025
  • 2 Oral Paper Presentation in IEEE IGARSS 2024
  • Oral Paper Presentation in BMVC 2023
  • Best Paper Award in IW-FCV 2022
  • Best Presentation Award in IW-FCV 2022
  • Outstanding Research CSE Thesis Award, 2021
  • Best Student Paper Award in IW-FCV 2020
  • Joint Secretary, PSAK, 2018-2019
  • Secretary Information, PSAK, 2017-2018
  • KNU International Graduate Scholarship (KINGS) for Ph.D. studies, 03.2016-02.2021
  • Fully funded scholarship by Sejong University for MS studies, 03.2014-02.2016
  • Fellowship by Pakistan Institute of Engineering and Applied Sciences (PIEAS) for BS (Computer and Information Science) studies, 2007-2011

Book Chapters & Patent

Book Chapter

Mustansar Fiaz, Arif Mahmood and Soon Ki Jung. "Deep Siamese Networks toward Robust Visual Tracking", ISBN 978-1-78985-158-8, IntechOpen, DOI: 10.5772/intechopen.86235.
Online

Patent

Mustansar Fiaz, Hisham Cholakkal, Sanath Narayan, Rao Muhammad Anwer, and Fahad Shahbaz Khan. "System and Method for Attention-aware Relation Mixer for Person Search", United States patent application US 17/983,741. 2024 May 9.