Parvez Ahmad

As a seasoned professional in the field of biometrics and computer vision, my career has been marked by a deep commitment to innovating and advancing the realms of deep learning and artificial intelligence. With a PhD background, I have extensive experience applying deep learning techniques to various domains, including object detection, human pose estimation, age prediction, multi-camera analysis, and advanced cancer segmentation in 2D and 3D medical imagery.

My expertise includes a strong proficiency in machine learning methodologies such as GANs, Autoencoders, Transformers, and Convolutional Neural Networks (CNNs), utilizing frameworks like TensorFlow, PyTorch, and Keras for machine learning projects. I am adept in leveraging OpenCV for real-time computer vision applications and have optimized model efficiency with TensorFlow Lite and PyTorch Lite, integrating these with microcontrollers for enhanced functionality.

Utilizing the Intel OpenVino toolkit, I have developed capabilities for real-time object detection and analysis, including age estimation and emotion analysis from video data and streams. My work is characterized by a deep understanding of programming languages and mathematics, essential for designing custom image processing algorithms for real-time applications and applying statistical analysis effectively in data-driven decision-making.

A significant project in my portfolio is the development of an AI-based Hawk-Eye system for cricket, using Python libraries like OpenCV and NumPy for frame-by-frame video analysis. This involved leveraging pre-trained deep learning models such as YOLO for ball tracking and player pose estimation, leading efforts in data collection, labeling, and automating the labeling process.

In addition to my technical expertise, I have been instrumental in enhancing and optimizing software applications, playing a pivotal role in the design and development of robust software systems, and developing databases using Oracle. My contributions have extended to the development of a time-sharing system to improve resource utilization. My professional journey is a testament to my passion for solving complex challenges and contributing to the advancement of computer vision and AI.

My educational journey, starting in 2016 as a Ph.D. student in Computer Applied Technology at Huazhong University of Science and Technology (HUST), Wuhan, China, has been centered around medical image segmentation, with a specific focus on stroke lesion segmentation. This specialization involved extensive use of the Python programming language, combined with skills in image processing, medical imaging, statistical analysis, and deep learning models.

During my undergraduate and postgraduate studies, I developed a strong foundation in various programming languages. However, it was during my Ph.D. studies that I significantly advanced my expertise in Python programming. Additionally, deep learning, a field not covered in my earlier academic pursuits, became a focal point of my Ph.D. research. I immersed myself in this area, gaining mastery, particularly in the application of deep learning to medical imaging.

The environment at HUST was instrumental in this development, providing a platform where I could both learn and apply deep learning techniques, not only in medical imaging but also in non-medical domains. This versatility of deep learning, spanning multiple fields, became a key insight from my academic journey.

My research has been driven by a goal to accelerate the diagnosis of brain cancer, aiming to improve patient outcomes. This led to my work on several medical datasets, developing and applying a singular deep-learning architecture across them. The culmination of my studies is marked by the publication of these research findings, demonstrating the practical impact of my work.

As I approach my graduation in June 2024, my Ph.D. studies, complemented by my employment history, highlight the integral role of deep learning across various domains, from medical to non-medical applications. This journey reflects a blend of academic rigor and practical application, underscoring the transformative potential of deep learning in diverse fields.