Daniel Angelov Email & Phone Number

Berkeley, Us

Efemarai is a platform for testing and robustifying Machine Learning models.

· Offering free consultations on understanding bias within your propitiatory dataset and ethical implications of developed ML algorithms.· Delivered full pipelines for systems using Machine Learning - defining the problem and expected outcome, data collection, annotation, to model building and designing deployment mechanism.· Worked with public and private sector companies to assess or deliver custom solutions for object detection, tracking, instance segmentation, optimization within the domain of asset management, contactless checkout, medical domain, sports industry etc.

Edinburgh, Gb

- Research Assistant on a project funded by the Alan Turing Institute and Cancer Research UK (CRUK) - Mammobot. Investigating uncertainty and confidence in multi-modal machine learning models (vision, pCLE, Raman spectroscopy, tabular) as part of a robot to aid early breast cancer detection.- Continued the DARPA XAI project at UoE by creating meta learning agents. - Tutoring Probabilistic Modelling and Reasoning, Reinforcement Learning and mentoring groups in System Design Projects.

Palo Alto, California, Us

Working on a project under DARPA’s explainable AI (XAI) program at PARC - COGLE (COmmon Ground Learning and Explanation). Aiming to build RL agents that not only solve a particular problem, but also give insight about why a specific action was taken.

Reading, Berkshire, Gb

· Working as a lab demonstrator for Part 1 Courses in Cybernetics,Robotics and Electrical Engineering.· Additionally volunteering as a Programming Classes Assistant for students to fully understand and practise the curriculum material.

Year 1: June - Sept 2013Year 2: May - Aug 2014· Mentors: Gary Bradski, Vincent Rabaud· I optimized and merged a new feature detection algorithm into the imgproc module of the OpenCVlibrary. The converted code ran in real time owing to being around 45% faster than the original.· Produced library standard quality code that was accepted into master/trunk.· Ongoing support and development.

· Created a lightweight C version of the roscpp stack that utilised BSD sockets and has nearly doubled the standard ROS node throughput performance.· Successfully running the node on a STM32 with FreeRTOS + lwIP in under 64 kB.· Bootstrapped a SaaS module on the Object Tracking system (prev. year) based on Machine Learning algorithms that has a 96.2% success rate on removing misclassification.· Mentoring a new intern in his duties for the project.

· Working in a small, motivated team, often under pressure due to the nature of the project.· My main task is related to Image Processing and Data Analysis. · Our team investigated the possibility of creating a real time stereo vision based system for identifying, positioning and tracking of flying morphable objects.· I am involved in building software for large scale systems, but also gaining experience in small embedded type systems.

Doctor Of Philosophy (Phd), Robotics And Autonomous Systems

Master Of Engineering (Meng), Robotics

Diploma Za Sredno Obrazovanie, Mathematics, Physics, Programming