sarali.bscs22seecs
About Candidate
I am Ahmad Sarmad Ali, a dedicated Computer Science student at NUST-SEECS (CGPA 3.23/4.0) with a passion for cutting-edge technologies in Computer Vision, Machine Learning, and Artificial Intelligence. With hands-on experience as a Deep Learning and Computer Vision Intern at SEECS and TUKL Lab, I have mastered tools like TensorFlow, Keras, Scikit-learn, CV2, Numpy, and Matplotlib, delivering innovative solutions across diverse domains.
My expertise spans object detection and localization (e.g., YOLO-based fruit detection), facial recognition/analysis, knowledge distillation, federated learning/unlearning, and AI applications in agriculture and optical networks. Notable achievements include developing a CNN-based autoencoder for fruit shelf-life detection, implementing federated unlearning for document analysis (conference paper in progress for ICDAR-2024), and creating early warning systems for flooding using sensor data. I’ve also contributed to IEEE Access journal papers and worked on real-time face filter applications at Walee.
Skilled in Python, Java, and MySQL, with certifications in Deep Learning and Problem Solving, I bring a blend of technical proficiency and research-driven innovation. Whether it’s 3D modeling from images, transformer-based car wheel detection, or Grad-CAM visualization for CNNs, I deliver high-quality, tailored solutions. Let’s collaborate to turn your ideas into reality—first impressions matter, and I’m here to exceed your expectations!
Salary
Job Categories ( For Example : Flutter Developer or Data Engineer Like That ) )
Location
Education
Work & Experience
Main Research topics: Early Warning Systems, Federated Learning, Knowledge Distillation, ML in optical Networking Worked on Quality of Transmission Estimation Using ML in Optical Networks. Implemented Knowledge Distillation Based QOT Estimation in Optical Networks. Developed a model to detect car wheels using sensor data. Explored Federated Learning in Optical Networks. Completed a journal paper related to federated learning based QOT estimation in IEEE Access journal
Main Research topics: Crops identification at different growth stages, Agri-Drone project, Shelf life detection Implemented a unique CNN based AUTOENCODER + SOFTMAX classifier for shelf life detection of fruits. Implemented YOLO based object detection model for fruits detection in crops. Worked on classifying crop filed images depending upon maturity level and then generating heat maps based on output Implemented algorithm to stitch images of field together taken from drone to generate full view of field.
Main Research topics: Machine Unlearning, Federated Unlearning, Asynchronous FL, Object Detection, Document analysis Explore different Machine unlearning techniques and implemented them. Implemented different Federated unlearning techniques and used them for unlearning of document data. Used YOLOv8 models in decentralized systems for detecting different entities in documents and them apply Federated unlearning to make model unlearn specific entity efficiently. Currently writing a conference paper on Federated Unlearning for document analyses for ICDAR-2024
Main Research topics: Face detection, recognition and landmark detection. Application of filters on human faces, Application of emotion based filters. 3D asset creation from images. Experimented different face detection and landmark detection models. Implemented human face landmark detection and apply filters on faces using those landmarks in real time. Experimented with Google’s mediapipe model for face detection and landmark detection. Working on face age progression filter module using Face Reaging GANs.