Konstantinos Peppas Email & Phone Number

Thessaloniki

Developed a 3D Human Pose Estimation from RGBD images implementation for Human-RobotCollaborative applications (Tensorflow, Python, OpenCV)Developed a real-time Human Activity Recognition implementation, which was integrated in amobile application aiming at changing users’ energy behavior (Keras, Python, Tflite, Java)Developed real-time Deep Learning.

Paris Area, France

Received intensive soft skills training ranging from presentation and consulting skills to teamworkand agile methodology workshopsDelivered presentations to an audience of more than 100 peopleExperienced working in a multicultural environment

Geneva Area, Switzerland

Thessaloniki, Greece

• Thesis ProjectDevelopment of methods and techniques of Distributed Deep Learning of Deep Neural Networks with Privacy Protection
• Research on state-of-the-art methods and techniques
• Development of an adaptable and extensible base code capable of simulating Distributed Deep Learning in the Keras framework with a Tensorflow back-end
• Development and implementation of four methods using the datasets MNIST and PAMAP2
• Using a combination of state-of-the-art methods, results comparable to a non-distributed implementation were achieved with a sufficient level of privacy protection
• The privacy protection intensive methods provided an attractive trade-off between privacy and performanceTools: Python, Keras, Tensorflow (back-end)

Thessaloniki, Greece

• Migration of the “front-end” from Qt and C++ to PyQt and Python, which accelerated the development process.
• Back-end development in Python for post-processing the product’s simulation results.
• Development and maintenance of utility python modules for internal purposes as well as on demand for clients, which facilitated the processes for the engineers.

Thessaloniki, Greece

Thessaloniki, Greece

• Developed a fully autonomous robot which won 2nd place in category autonomy in RoboCup Rescue 2015 in Hefei China
• Implemented the agent of the robot as a generic finite state machine which simplified the process of adding and removing nodes (on Python)
• Established an automated testing framework that made the process of writing functional tests easier and more effective (on Python)
• Fixed and refactored the Sensors' Controller (on Python and C++)
• Implemented unit and functional testing on the Agent and Sensors' Controller nodes