Musaib P. - Machine Learning Lab – Autonomous Driving & Generative Models

Go to Website

Erlangen, Germany

Experience

Apr 2025 - Aug 2025

5 months

Erlangen, Germany

Machine Learning Lab – Autonomous Driving & Generative Models

Friedrich-Alexander-Universität Erlangen-Nürnberg

Built an ML pipeline for autonomous driving in OpenAI Gymnasium, generating labeled datasets from simulation frames and control actions.
Designed and trained lightweight CNN and RNN models for end-to-end control and implemented behavioral cloning with model compression for real-time inference.
Designed and trained a diffusion model from scratch on MNIST, implementing denoising, sampling, and image generation during inference.

Jan 2025 - Mar 2025

3 months

Erlangen, Germany

Canny Edge Detection Optimization (DSP Lab Project)

Friedrich-Alexander-Universität Erlangen-Nürnberg

Implemented the full Canny Edge Detection pipeline in Python using OpenCV, modularizing key stages such as gradient computation and hysteresis thresholding.
Enhanced performance by integrating CLAHE preprocessing and multi-scale Gaussian filtering, improving edge contrast and localization.
Conducted experiments on the BSDS500 dataset to analyze the impact of parameter tuning and multi-scale edge combination.

Sep 2022 - May 2023

9 months

Aligarh, India

Undergraduate Thesis – EEG Seizure Detection Using CNN

Aligarh Muslim University

Developed a CNN-based model for epileptic seizure detection using CHB-MIT EEG data.
Compared time-domain and FFT-based features, achieving superior performance with frequency-domain inputs.
Designed a lightweight CNN for real-time classification of seizure states (Interictal, Preictal, Ictal).

Summary

M.Sc. student majoring in Machine learning in Signal Processing with a strong foundation in machine learning, signal processing, and deep generative models.

Experienced in implementing neural architectures for perception and generation tasks.

Seeking a Master’s thesis or working student position in AI, focusing on deep learning, generative modeling, or intelligent systems.