About me

I am currently in my 4th year of pursuing a Ph.D. under the supervision of Dr. Vishal Patel at the Vision and Image Understanding (VIU) Lab in the Department of Electrical and Computer Engineering at Johns Hopkins University. My research encompasses the realms of computer vision and deep learning, primarily concentrating on investigating and mitigating data memorization issues, as well as addressing fairness and bias in generative models and vision-language models (VLMs). I am also interested in applying deep generative models to low-level vision tasks, particularly in image restoration.

I graduated with a Bachelor’s degree in Biomedical Engineering from University of Moratuwa, Sri Lanka in 2020, where I completed my thesis on designing a sEMG sensor system for controlling a bionic hand, advised by Dr. Simon Kappel.

I worked as a Research Assistant at Agency for Science Technology and Research (A*STAR), Singapore under the supervision of Dr.Bhanu Prakash and Dr. Shigeki Sugii.

Recent News

• Oct 2024 - “Frame by Familiar Frame: Understanding Replication in Video Diffusion Models” was accepted to WACV 2025.
• Oct 2024 - Participated as a Panelist in the Amazon Science and Internship Discussion at the AI2AI Research Festival.
• May 2024 - Started my summer internship at Amazon as an Applied Scientist Intern.
• Sep 2023 - Paper on Analyzing Bias in Diffusion-based Face Generation Models accepted at IEEE International Joint Conference on Biometrics (IJCB) 2023.
• March 2023 - Paper on SAR Despeckling using a Denoising Diffusion Probabilistic Model accepted to IEEE Geoscience and Remote Sensing Letters
• June 2022 - Attended CVPR 2022, New Orleans
• April 2022 - Two papers accepted at International Geoscience and Remote Sensing Symposium 2022 IGARSS
• August 2021 - Started my PhD at Johns Hopkins University

Selected Publications

• Frame by Familiar Frame: Understanding Replication in Video Diffusion Models
  A. Rahman*, M. V. Perera*, and V. M. Patel. IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2025 [paper]

• Analyzing Bias in Diffusion-based Face Generation Models
  M. V. Perera, and V. M. Patel. IEEE International Joint Conference on Biometrics (IJCB) 2023 [paper]

• SAR Despeckling using a Denoising Diffusion Probabilistic Model
  M. V. Perera, N. G. Nair, W. G. C. Bandara, and V. M. Patel. IEEE Geoscience and Remote Sensing Letters 2023 [paper], [code]

• Transformer-based SAR Image Despeckling
  M. V. Perera, W. G. C. Bandara, J. M. J. Valanarasu, and V. M. Patel. IEEE International Geoscience and Remote Sensing Symposium (IGARSS) 2022 [paper], [code]

• SAR Despeckling using Overcomplete Convolutional Networks
  M. V. Perera, W. G. C. Bandara, J. M. J. Valanarasu, and V. M. Patel. IEEE International Geoscience and Remote Sensing Symposium (IGARSS) 2022 [paper], [code]

• A Joint Convolutional and Spatial Quad-Directional LSTM Network for Phase Unwrapping
  M. V. Perera*, A. De Silva*. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2021. [pdf]

• Real-Time Hand Gesture Recognition Using Temporal Muscle Activation Maps of Multi-Channel SEMG Signals
  A. De Silva*, M. V. Perera*, K. Wickramasinghe, A. M. Naim , T. D. Lalitharatne and S. L. Kappel. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Barcelona, Spain, 2020. [pdf]

[* denotes equal contribution]

Full list of publications can be found here.