Muhammad Mubeen

Islamabad · Pakistan · 44090 · +92-333-565-1195 · [email protected]

Hey, I'm Muhammad Mubeen, an experienced Machine Learning and Computer Vision Engineer with 4+ years of relevant experience. Passionate about staying on the cutting edge of technology, I'm dedicated to using my expertise to develop innovative solutions and optimize systems for maximum efficiency. Whether it's collaborating with others or driving positive change in the industry, my goal is always to bring value to every project.

Experience

Senior Machine Learning and Computer Vision Engineer

Darvis

Worked as a software engineer at a US-based startup focused on automating the tracking of various items in different environments.

  • Developed and implemented various features such as face anonymization and plug-and-play model support for Nvidia's Deepstream-based pipelines.
  • Designed and executed optimized end-to-end vision pipelines using GStreamer/Deepstream for different hardware platforms, including Nvidia GPUs and OpenVINO.
  • Created an AI-powered counting algorithm with an accuracy rate of over > 80% for item insertions in complex environment.
  • Developed AI-based applications for gun detection in uncontrolled environments and bed state tracking for hospitals, available on android-based devices (AzenaOS).
February 2022 - Present

Machine Learning Engineer - II

Veeve

Worked as part of the research team at a US-based start-up focused on developing a self-checkout shopping cart.

  • Contributed to the productionization and optimization of various algorithms in C/C++ for Nvidia's Jetson devices, using the Nvidia DeepStream SDK and GStreamer framework for efficient inference pipelines.
  • Leveraged LSVR and Few-Shot Learning techniques to enhance the accuracy of SKUs/PLUs detection, resulting in up to 10% model accuracy improvement and 3x inference speedups.
  • Implemented depth stream compression to minimize storage requirements, resulting in a more than 10x reduction in storage space.
  • Developed a shrinkage detection system to track the items inserted into the shopping cart, reducing the incidence of theft.
January 2020 - January 2022

Machine Learning Engineer

OPTIMAL Lab

Developed a portable Edge AI system to assist the visually impaired by detecting and categorizing indoor and outdoor items, estimating their relative distances, and providing audio guidance.

  • Designed and implemented the various components of the system, including the input stream, inference, and output feedback, using C/C++.
  • Conducted data collection, trained and tested models, and optimized the model inference to achieve up to 3x speed-ups.
  • Validated the final prototype with visually impaired individuals, receiving positive feedback on the system's design and usability.
July 2018 - July 2019

Research Engineer

OPTIMAL Lab

As a research assistant at Optimal Lab, I worked on my final year project and contributed to the advancement of the lab's research initiatives.

  • The aim of this final year project was to train deep learning based models to extract speaker-specific information from speech signal invariant to the environment, noise, and other factors.
  • Various approaches including Auto-Encoders, Contrastive Learning, Siamese Networks etc. were used to extract information from MFEC based speech features and then clustered using Gaussian Mixture Models.
  • In the end, we achieved near state of the art results on NIST based datasets (TIMIT, NTIMIT, KING)
August 2017 - June 2018

Education

National University of Sciences & Technology

Bachelor of Electrical Engineering
Machine Learning - Computer Vision - Digital Image Processing - Digital Signal Processing

GPA: 3.02

September 2014 - June 2018

Skills

Programming Languages
Technologies
  • Nvidia DeepStream, Intel OpenVino, GStreamer
  • PyTorch, Tensorflow, TensorRT, CAFFE, Keras, Sklearn
  • Nvidia Jetson Devices
  • OpenCV, Numpy, Pandas, Scipy, Matplotlib
  • Docker, VS Code, Git, GCP, AWS

Interests

Apart from being a software developer, I am a big anime fan. With favorites like My Hero Academia, Naruto, Dragon Ball Z, and Full Metal Alchemist, I love to watch them and participate in the larger anime community.

When forced indoors, I follow a number of sci-fi and fantasy genre movies and television shows, I am an aspiring chef, and I spend a large amount of my free time exploring the latest technology advancements in the front-end web development world.

Download My CV

Click here to view my resume.