Reza AzadDeep Learning Researcher & founder of Bmdeep
LFB RWTH |
 |
Biography
I am a self-motivated and goal-oriented deep learning researcher on the intersection of machine learning, computer vision and medical image analysis. My ultimate research purpose is to mimic human intelligence and design intelligent algorithms to cope with challenges in the medical domain. I have a primary focus on the following topics:
1- Few-shot learning for medical image analysis
2- Texture and Inductive bias in CNN networks
3- Transformers for medical image segmentation
4- Minimizing human supervision (self-supervised, semi-supervised and multi-modality learning) along with robust algorithms to tackle the problem of missing labels, modalities and imperfect medical data.
5- Read-world challenges for medical image analysis including Kaggle and Grand-challenge
Before joining RWTH, I was a research internship at MILA/NeuroPoly and ETS University, Montreal , working with Prof. Jose Dolz. I obtained my M.Sc. degree under supervision of Professor Shohreh Kasaei, Sharif University of Technology in September 2017.
I am always open to research collaboration. So if you are interested in healthcare, machine learning, computer vision, and medical image processing, feel free to drop me an email with your CV.
Experience
-
RWTH University, Aachen, Germany2021-2022
Research Fellow
Topic: Medical Image Analysis using Deep Learning -
MILA/NeuroPoly (Google brain), deep learning for medical imaging research group, Montreal, Canada2020 – 2021
Applied Research Intern
Topic: Intervertebral Disc Labeling using Deep Learning
-
ETS, Montreal, Canada2019. 2020
Research Intern
Topic: Few-shot learning
-
Sharif University of Technology, Tehran, Iran2015. 2017
Master of AI
Topic: Human Action Recognition
Selected Publications [Google Scholar]
|
Contextual Attention Network: Transformer Meets U-Net
|
|
SMU-Net: Style matching U-Net for brain tumor segmentation with missing modalities
|
|
Medical Image Segmentation on MRI Images with Missing Modalities: A Review
|
|
Deep Frequency Re-Calibration U-Net for Medical Image Segmentatio
|
|
Stacked Hourglass Network with a Multi-level Attention Mechanism: Where to Look for Intervertebral Disc Labeling
|
|
Multi-scale Regional Attention Deeplab3+: Multiple Myeloma Plasma Cells Segmentation in Microscopic Images
|
|
On the texture bias for few-shot cnn segmentation
|
|
Attention Deeplabv3+: Multi-level Context Attention Mechanism for Skin Lesion Segmentation
|
|
Semi-supervised few-shot learning for medical image segmentation
|
|
Multi-level Context Gating of Embedded Collective Knowledge for Medical Image Segmentation
|
|
Bi-Directional ConvLSTM U-Net with Densely Connected Convolutions
|
|
Dynamic 3D Hand Gesture Recognition by Learning Weighted Depth Motion Maps
|
Honors & Awards
| IEEE journal reviewer, 2021 |
| Second place in SegPC grand challenge. SegPC 2021 ISBI Challenge |
| Invited speaker at the fourth IPM advanced school on computing 2020. IPM event |