Akbar Ali

restores what others discard.

Research focus
Computer Vision · Generative Models · Hyperspectral Imaging · Latent Inversion · Gaussian Splatting
Download CV View publications Memories
Scroll to explore

Vision meets
the hard data.

Akbar Ali

IIT Gandhinagar — Helmholtz Freiberg

I am a fourth-year Ph.D. student in Computer Science and Engineering at IIT Gandhinagar, advised by Prof. Shanmuganathan Raman. I lead an independent research line on generative priors for image restoration — latent inversion in StyleGAN and diffusion models, zero-shot and training-free restoration, and 2D/3D Gaussian splatting for hyperspectral imaging.

My work sits deliberately across disciplines. I collaborate with computer scientists, earth scientists and agricultural researchers, and I have built and released datasets that serve all three. I am currently a Visiting Researcher at the Helmholtz Institute Freiberg (HZDR), Germany, on the Research Lab Abroad Fellowship under the Visvesvaraya PhD Scheme (MeitY), working on hyperspectral restoration with generative models.

Before IIT Gandhinagar, I was a Project Associate at IIT Ropar on hyperspectral-stereo fusion for industrial volume estimation. I hold an M.Tech (2nd rank) and B.Tech (1st rank) in CSE from Aliah University, Kolkata, and a UAV Pilot Licence for platforms up to 25 kg.

I am open to research positions at the boundary of visual learning, generative modelling and real-world sensing. Get in touch.

Updates

What's
happening.

Apr 2026 GenR accepted in Pattern Recognition Letters (Elsevier).
Mar 2026 Completed six months at Helmholtz Institute Freiberg (HZDR), Germany.
Sep 2025 Presented two papers at IEEE ICIP 2025, Anchorage, Alaska.
Jul 2025 Both ICIP papers published on IEEE Xplore.
May 2025 Awarded Research Lab Abroad Fellowship (Visvesvaraya PhD Scheme, MeitY) for placement at HZDR.
May 2025 Two papers accepted at IEEE ICIP 2025 (COT-AD and IRUSIR).
Jun 2024 Attended 3D Vision Summer School, IIIT Bangalore.
May 2024 Presented at the Visvesvaraya PhD Scheme Workshop, IIT Mandi.
May 2023 First field campaign with the Earth Science group.
Early 2023 Certified as a UAV Pilot for platforms up to 25 kg.
Feb 2023 Parkinson's disease paper published by Springer Nature.
Dec 2022 Joined IIT Gandhinagar as a Ph.D. scholar.
Selected output

Pixels degraded.
Ideas restored.

2026
GenR
Pattern Recognition Letters — Journal Published
GenR: Robust Generative Face Image Restoration in Latent Space
Akbar Ali, Indra Deep Mastan, Shanmuganathan Raman.
A degradation-agnostic face restoration method that optimises directly in the latent space of a pretrained generator, handling denoising, deblurring and inpainting without retraining per degradation.
paper code
2026
Deblurring
Under review, 2026
Unsupervised Image Deblurring in Latent Space using Pivotal Tuning Inversion
Akbar Ali, Indra Deep Mastan, Shanmuganathan Raman.
Pivotal tuning inversion anchors a blurred observation to a generator's latent manifold, recovering sharp detail without paired supervision or a known blur kernel.
2026
ZS-CLIP2Clean
Under review, 2026
ZS-CLIP2Clean: Zero-Shot CLIP-Guided Training-Free Single-Image Denoising
Abhiyodaya Pandey, Akbar Ali, Indra Deep Mastan, Shanmuganathan Raman.
Denoising a single image with no training and no clean reference, using CLIP as a semantic critic to steer restoration towards a perceptually faithful result.
2026
InscriptionOCR
Under review, 2026
InscriptionOCR: A Dataset and Method for Understanding Inscriptions
Jaidev Sanjay Khalane, Akbar Ali, V. N. Prabhakar, Shanmuganathan Raman.
A dataset and recognition pipeline for historical stone inscriptions, where weathering, low contrast and archaic glyph forms break standard scene-text OCR.
2025
IRUSIR
IEEE ICIP 2025 — Conference
Investigating Robustness of Unsupervised StyleGAN Image Restoration
Akbar Ali, Indra Deep Mastan, Shanmuganathan Raman.
A stress test of unsupervised StyleGAN restoration under degradations it was never tuned for, showing where latent-space optimisation holds up and where it silently fails.
paper data
2025
COT-AD
IEEE ICIP 2025 — Conference
COT-AD: Cotton Analysis Dataset
Akbar Ali, Pankaj Khanna, Subramanian Sankaranarayanan, Shanmuganathan Raman.
An aerial and ground image dataset covering a full cotton growth cycle, collected over weekly drone missions at 10 m, 15 m and 115 m from July to December 2023, for crop health and pest analysis.
paper data
2023
Parkinson's
Springer Nature, 2023
Descriptive Predictive Model for Parkinson's Disease Analysis
Akbar Ali, Ranjit Kumar Rout, Saiyed Umer.
A predictive framework for Parkinson's disease analysis pairing classification accuracy with descriptive feature analysis, so clinicians can see which measurements drive each prediction.
paper
Background

The path
so far.

Mar — Sep 2026
Freiberg, Germany
Visiting Researcher — Helmholtz-Zentrum Dresden-Rossendorf
Helmholtz Institute Freiberg · Supervised by Prof. Dr. Richard Gloaguen · mentored by Dr. Sam T. Thiele
Research Lab Abroad Fellowship (Visvesvaraya PhD Scheme, MeitY). Worked in the Exploration department on hyperspectral image restoration with generative models, improving spectral data quality for more accurate downstream analysis.
Dec 2022 — present
Gandhinagar, India
Ph.D., Computer Science and Engineering — CPI 9.14/10
IIT Gandhinagar · advised by Prof. Shanmuganathan Raman
Working on visual learning and generative representations, with the aim of building interpretable and efficient systems for image restoration and hyperspectral imaging.
Sep 2021 — Nov 2022
Ropar, India
Project Associate — IIT Ropar
IIT Ropar · with Prof. Neeraj Goel
Volume estimation of complex objects using hyperspectral imaging and stereo vision fusion for industrial automation and robotics. First sustained work at the boundary of sensing hardware and vision algorithms.
Aug 2019 — Jul 2021
Kolkata, India
M.Tech, Computer Science and Engineering — CPI 9.03/10 · 2nd rank
Aliah University · with Prof. Saiyed Umar
Thesis on a descriptive predictive framework for Parkinson's disease analysis, pairing classification accuracy with interpretable feature analysis. Published by Springer Nature in 2023.
Jun 2015 — Jul 2019
Kolkata, India
B.Tech, Computer Science and Engineering — CPI 8.97/10 · 1st rank
Aliah University
IoT-based home automation system for remote control and monitoring using sensor integration and wireless communication. This project is what turned my interest towards machine learning and deep learning.
Teaching and mentoring

Passing it
forward.

Jan — Apr 2025
Lead Teaching Assistant — ES 113, Data Centric Computing
IIT Gandhinagar
Led the TA team for 45 students, covering data structures, algorithms, data analysis in Python (Pandas, Matplotlib), SQL and introductory machine learning. Set the QA session format and coordinated grading standards across assistants.
Aug — Nov 2024
Graduate Teaching Fellow — FP 602, Writing
IIT Gandhinagar
Guided 35 students through classwork, assessment and progress tracking, holding individual reviews with students falling behind.
June 2024
Teaching Assistant — SC 333, Drone Data Acquisition and Processing
IIT Gandhinagar
Ran flight training and field sessions for around 30 students on drone data acquisition and processing with RGB and multispectral imagery. Responsible for pre-flight checks, airspace discipline and student safety during live flights.
Dec 2022 — Apr 2024
Teaching Assistant — ES 114, Probability, Statistics and Data Visualization
IIT Gandhinagar
Ran labs and tutorials for 50+ students across three semesters using the Python data science stack, and supported the Probability and Randomization course.
Ongoing
Student Mentoring
IIT Gandhinagar
Supervise B.Tech and M.Tech students on hyperspectral imaging, image editing and GAN inversion projects. Two of these collaborations became co-authored submissions.
Collaborations and field leadership

Where the
data begins.

Feb 2024 — Jan 2025
3D Geophysical Image Translation using GANs
A year-long collaboration with geoscientists, translating 3D geophysical scans into photorealistic virtual outcrop geology. Required learning enough field geology to evaluate what a plausible result looks like, then translating that back into a generative formulation.
Jun 2023 — Jan 2024
Smart farming of cotton using aerial imagery — COT-AD
Designed and ran a six-month acquisition campaign with agricultural researchers, flying weekly missions at 10 m, 15 m and 115 m to track weather and pest effects across a full growth cycle. I owned the flight schedule, the multi-altitude protocol, and the dataset that became COT-AD.
Cotton growth monitoring
2023 — 2024
Drone and core data acquisition — Earth Science, IIT Gandhinagar
Planned and led three multi-day campaigns (May 2023, Aug 2024, Nov 2024) for aerial and core sample collection, coordinating logistics, equipment and flight operations. Site names withheld for confidentiality.
Field campaign Field campaign
Jun 2024
3D Vision Summer School — IIIT Bangalore
Week-long school on deep learning for 3D vision. The Gaussian splatting sessions here shaped the direction my hyperspectral work took afterwards.
3D Vision Summer School 3D Vision Summer School
May 2024
Visvesvaraya PhD Scheme Workshop — IIT Mandi
Presented my work to an interdisciplinary audience of scholars from across engineering disciplines. Explaining your work to people outside your field is the fastest way to find out which parts you actually understand.
IIT Mandi workshop IIT Mandi workshop
Awards and certifications

Recognition
along the way.

2023
UAV Pilot Licence
Certified for platforms up to 25 kg. DJI Mavic and Bajrang platforms. Mission planning, spectral capture, aerial inspection.
2022
UGC-NET
Computer Applications.
2021
2nd rank overall, M.Tech
Aliah University, Kolkata. CPI 9.03/10.
2019
1st rank overall, B.Tech
Aliah University, Kolkata. CPI 8.97/10. Also 1st rank in AUAT M.Tech entrance.
2015 — 21
National Scholarship — five times
Three during B.Tech, two during M.Tech.
2015
64th rank, AUAT B.Tech entrance
Also qualified BCECE, WBJEE and AIEEE.

Good work
starts here.

Interested in a postdoc, research collaboration, or just want to talk about generative priors and hyperspectral data? I would love to hear from you.

akbar.ali@iitgn.ac.in Google Scholar github.com/aamaanakbar LinkedIn

I formulate research problems end to end and take them from idea to publication. I work across disciplines and I am comfortable in the field as well as in the lab.

Current interests: hyperspectral restoration, training-free image recovery, 3DGS for non-RGB imaging, generative models for scientific data.

Download full CV

Akbar Ali — aamaanakbar.github.io — last updated September 2026
Back to top Memories Email