Meshkat Ahmad

Yangpu District, Shanghai, PR China

meshkat22@m.fudan.edu.cn

ahmadmeshkat26@gmail.com

I am a graduate student at Fudan University working on brain–computer interfaces (BCI). My M.Eng. in Electronic Information finishes in December 2026, and most of my research so far has been about making sense of noisy EEG and EMG signals — figuring out what a person intends to do with their hand, and turning that into something a device can use.

These days I am at NeuroRevix in Shanghai as an algorithm intern. The team is building a hand rehabilitation system for post-stroke patients, and my part is the EEG+EMG decoding: collecting data, cleaning it up, training models, and checking how well they hold up in practice.

My first paper, MINDGAN, came out of my work at Fudan. It pairs a GAN-based data augmentation step with a hybrid CNN–Transformer classifier. It was published at IEEE ICCET 2026 and also won the Best Presenter Award. The code is on GitHub if you want to look inside.

Things I keep coming back to:

motor imagery and movement intention decoding from EEG/EMG
CNN–Transformer architectures for biosignals
data augmentation for the small, messy datasets that biomedical work actually gives you
real-time BCI and rehabilitation applications

I am looking for a PhD position or a research role in BCI, neural signal processing, or applied deep learning, starting around early 2027. If any of this sounds relevant, write to me at meshkat22@m.fudan.edu.cn or ahmadmeshkat26@gmail.com.

news

Aug 12, 2026	Started as an algorithm intern at NeuroRevix (Shanghai) — working on EEG+EMG decoding for a post-stroke hand rehabilitation system.
Jun 20, 2026	Our paper MINDGAN is out on IEEE Xplore (ICCET 2026, Guangzhou), and it picked up the Best Presenter Award. Code
May 10, 2026	Released MINDGAN BCI Suite — a desktop app that covers the whole motor-imagery BCI loop: real-time inference (LSL & Cortex API), training, and drone control.
Jan 15, 2026	Built a small BCI game with the Emotiv EpocX — tug of war, played with mental commands. Demo
Mar 20, 2025	Finished the face-tracking drone project — OpenCV and MediaPipe driving a Tello. Demo

selected publications

ICCET
MINDGAN: EEG-Based Motor Imagery Decoding via Hybrid CNN–Transformer with Curriculum GAN Augmentation

Meshkat Ahmad, Yanqi Huang, and Xiaomei Wu

In 2026 9th IEEE International Conference on Communication Engineering and Technology (ICCET), Jun 2026

Best Presenter Award Abs DOI Bib HTML Code

Best Presenter Award

Decoding motor imagery from non-invasive EEG remains challenging due to the low signal-to-noise ratio and scarce labeled data across subjects. We present MINDGAN, a unified framework that integrates class-conditional deep convolutional GAN (cDCGAN) augmentation with a hybrid CNN–Transformer classifier under a three-phase curriculum training scheme supported by a quality-filtered replay buffer. A systematic six-configuration ablation reveals that recurrent components (LSTM) are detrimental within hybrid architectures. MINDGAN achieves competitive performance — 81.17% accuracy on BCI Competition IV Dataset 2A and 87.16% on Dataset 2B — with the lowest cross-subject variance on 2A.
@inproceedings{ahmad2026mindgan, title = {MINDGAN: EEG-Based Motor Imagery Decoding via Hybrid CNN--Transformer with Curriculum GAN Augmentation}, author = {Ahmad, Meshkat and Huang, Yanqi and Wu, Xiaomei}, booktitle = {2026 9th IEEE International Conference on Communication Engineering and Technology (ICCET)}, year = {2026}, month = jun, address = {Guangzhou, China}, pages = {18--24}, publisher = {IEEE}, doi = {10.1109/ICCET70051.2026.11659043}, url = {https://doi.org/10.1109/ICCET70051.2026.11659043}, }