Biography & Research Interest

Mehdi Sefidgar Dilmaghani is second year PhD student working in the field of computer vision. Mehdi started his work on signals since his Master’s studies when his main focus was on signal and image processing. Mehdi has studied both of his BSc and MSc in Iranian universities in the field of Electronics Engineering and because of the emphasize on mathematical courses in Iranian universities for engineering students he got skills in circuit design and math at the same time. To be able to combine both of these skills he chose electronic implementation of signal processing algorithms as his main field of interest. His BSc project was a simple voice detector and his MSc thesis was based on second generation of wavelets. 

Research Activities

My main focus is Smart Augmentation and GAN based methods. These methods and their outputs are of great importance for two main reasons:
They tackle the problem of data shortage in data thirsty networks.
Since these data are fake, GDPR and other limiting issues can not prevent their usage in different applications.
I started with writing two different networks in PyTorch (one classifier and one autoencoder), then I was trying to make those two networks to work in pipeline mode together and there were some challenges in that phase, however it was a very important step because I will have to do the same job after I finished the smart augmentation and switched to the GANs. Currently my main focus is to optimize this hybrid networks and get better results compared to the previous work done in our group. In the next level I will start my works with GANs, and the experience of this level will be helpful in the whole process.

Journal Publications

Shariff W, Dilmaghani MS, Kielty P, Lemley J, Farooq MA, Khan F, Corcoran P. Neuromorphic Driver Monitoring Systems: A Computationally Efficient Proof-of-Concept for Driver Distraction Detection. IEEE Open Journal of Vehicular Technology. 2023 Oct 18. 

Kielty P, Dilmaghani MS, Shariff W, Ryan C, Lemley J, Corcoran P. Neuromorphic Driver Monitoring Systems: A Proof-of-Concept for Yawn Detection and Seatbelt State Detection using an Event Camera. IEEE Access. 2023 Sep 5. 

S. Azizian, V. A. Aghchegala, S. Azizian, M. S. Dilmaghani. CMOS Bulk-Controlled Fully Programmable Neuron for Artificial Neural Networks, IETE Journal of Research (2018), https://doi.org/10.1080/03772063.2018.1431570

S. Azizian, V.Aziz Aghchegala, M. S. Dilmaghani, E. Mahoor, "CMOS Low-Power, FullyProgrammable Gaussian/Trapezoidal Fuzzy Membership Function Generator", Submitted in Transactions on Design Automation of Electronic Systems

Regular Conference Publications

Waseem Shariff, Mehdi SefidgarDilmaghani, Paul Kielty, & Peter Corcoran. (2023, September 2). Sensing Without Seeing - Event Camera Privacy. Irish Machine Vision and Image Processing Conference 2023 (IMVIP2023), University of Galway, Ireland. Zenodo. https://doi.org/10.5281/zenodo.8310913 

Paul Kielty, Cian Ryan, Mehdi SefidgarDilmaghani, Waseem Shariff, Joe Lemley, & Peter Corcoran. (2023, August 8). Neuromorphic Seatbelt State Detection for In-Cabin Monitoring with Event Cameras. Irish Machine Vision and Image Processing Conference 2023 (IMVIP2023), University of Galway, Ireland. https://doi.org/10.5281/zenodo.8223905 

Muhammad Ali Farooq, Waseem Shariff, Mehdi SefidgarDilmaghani, Wang Yao, Moazam Soomro, & Peter Corcoran. (2023, July 18). Decisive Data using Multi-Modality Optical Sensors for Advanced Vehicular Systems. Irish Machine Vision and Image Processing Conference 2023 (IMVIP2023), University of Galway, Ireland. https://doi.org/10.5281/zenodo.8160053 

Dilmaghani MS, Shariff W, Ryan C, Lemley J, Corcoran P. Control and evaluation of event cameras output sharpness via bias.InFifteenth International Conference on Machine Vision (ICMV 2022) 2023 Jun 7 (Vol. 12701, pp. 455-462). SPIE. 

Kielty P, Dilmaghani MS, Ryan C, Lemley J, Corcoran P. Neuromorphic sensing for yawn detection in driver drowsiness.InFifteenth International Conference on Machine Vision (ICMV 2022) 2023 Jun 7 (Vol. 12701, pp. 287-294). SPIE. 

B.Moradkhani, M. S.Dilmaghani, A. B. Rezaei. ANewApproachtoOptimizetheDeffuzification Unit of Fuzzy Systems, 27th Iranian Conference on Electrical engineering (ICEE2019), Yazd, Iran, DOI: 10.1109/IranianCEE.2019.8786479

B.Moradkhani,A.B. Rezaei, M. S.Dilmaghani. MergingDACAndFuzzifierStructures Asa New Approach in Membership Function Generators, 26th Iranian Conference on Electrical engineering (ICEE2018), Mashhad, Iran, DOI: 10.1109/ICEE.2018.8472409

M. S. Dilmaghani, D. Asemani. Correction of TimingSkew Error in TI-ADC’s Using Second Generation Wavelets, 25th Iranian Conference on Electrical engineering (ICEE2017), Tehran, Iran, DOI: 10.1109/IranianCEE.2017.7985493

M. S.Dilmaghani,S.Daneshvar,M. Dousty. AnewMRI andPETimagefusionalgorithmbased on BEMD and IHS methods, 25th Iranian Conference on Electrical engineering (ICEE2017), Tehran, Iran, DOI: 10.1109/IranianCEE.2017.7985242

Conference Participation & Presentations

27th Iranian Conference on Electrical Engineering (ICEE 2019), Yazd,Iran

26thIranianConferenceonElectricalEngineering(ICEE2018),Mashhad,Iran

25thIranianConferenceonElectricalEngineering(ICEE2017),Tehran,Iran