FIRSTNAME LASTNAME Tel Location Email LinkedIn URL An experienced Computer Vision, Machine Learning Engineer with success in leading and implementing projects on re-training and deploying deep learning models. Applies problem solving and innovation to adopt the latest quantization and sparsity techniques to optimise model speed and power consumption, while preserving state of the art inference accuracy and presented results to key stakeholders. Re-trained models for classification, object detection and instance segmentation using Tensorflow and Keras with strong attention to detail in underfitting and overfitting. TECHNICAL SKILLS Python SQL Keras OpenCV MATLAB Bash Tensorflow Detectron Data Cleaning Office 365 PROFESSIONAL EXPERIENCE Computer Vision / Machine Learning Engineer, Company X 2016 - Present Created and deployed Deep Learning models on our specially designed hardware, aimed at inference on the edge. ● Created quantized models, 8-bit or 4-bit activations and weights, which were 75% smaller than FP32 models, thus saving memory. Achieved 1000 Frames Per Second, with only a loss of 1% in accuracy on the Imagenet dataset, consumed less power during inference, still yielding high test accuracy. ● Led a team to compete in the IEEE LPIRC challenge. Organised project using Gantt Charts and Kanban Boards which allowed reallocation of resources to the critical path based on the slack time of other tasks. Overcame all obstacles with effective team communication and tackled early signs of failure. ● Re-trained object detection models including ResNet-SSD and MobileNet-SSD using Tensorflow on newly created datasets. Utilised Detectron2 for instance segmentation with Mask-RCNN models. Created fewer layer classification models on Keras to avoid underfitting on smaller specific datasets. ● Explored the trade-offs with image classification models in terms of accuracy and performance. Compiled metric data in databases - queried using SQL - and presented results to management and customers to verify product specifications. ● Contributed to the Intel Incubator Program, aimed at helping startups to re-train and deploy deep learning models. Maintained regular contact throughout the program to help the companies meet their goals and objectives, gave advice on the direction for the product to progress. ● Coordinated department’s internship program. Liaised with university programs, obtained funding for hiring and conducted interviews. Created and supervised projects on computer vision and IoT and mentored students. Name Surname Email Page Two EDUCATION ME in Electronic and Computer Engineering - First Class Honours, University Name - 2006 Research: Optimised algorithms for Electrostatic Kinetic Energy Harvesters (EKEH). Created mathematical models of the behavior of capacitors using MATLAB and processed simulated data of power generation and energy harvesting. Performed data analysis to obtain critical points in the system to maximize and optimise the harnessed electricity output. Published a research paper (below) to an IEEE conference and collaborated with a research team in France to verify findings with physical devices. B. Sc. in Electronic and Computer Engineering - Honours 2.1, University Name PUBLICATIONS Add here.