Developed and optimized computer vision models for deployment on edge devices.Employed techniques such as quantization, pruning, and model compression.Collaborated with teams to integrate models into edge systems.Utilized tools like Python, TensorFlow, PyTorch, ONNX, and OpenCV.

Image Retrieval Model for Architectural Engineering ApplicationsI developed a computer vision-based image retrieval model designed for architectural engineering applications. The project focused on retrieving similar images to an input hand-drawn geometrical shape from a dataset of architectural sketches. The retrieved images are used in a larger pipeline, where they are converted into CAD formats.Key contributions included:Designing and training a neural network model capable of efficiently retrieving visually similar hand-drawn shapes from the dataset.Implementing data augmentation techniques to enhance model performance.Fine-tuning the model to improve retrieval accuracy and speed for real-world architectural scenarios.Collaborating on integrating the model into a software pipeline for seamless CAD conversion of retrieved images.

Developed and Implemented a water meter reading app for a water company using C/C++ and QT.Designed and optimized user interfaces and backend logic.Collaborated with team members to integrate application features.Conducted testing and debugging to ensure software reliability.Documented development processes and provided support for end-users