This three day short course and workshop provides an in-depth presentation of programming tools and techniques for various computer vision and deep learning problems encountered in drone imaging. Special attention will be paid to drone cinematography, which is one of the main application areas of drone technologies. The same machine learning and computer vision problems do occur in other drone applications as well, e.g., for land/marine surveillance, search&rescue, infrastructure/building inspection and 3D modeling. The short course consists of three parts (A,B,C), each having lectures and a programming workshop with hands-on lab exercises.
Part A will focus on Deep Learning. The lectures of this part provide a solid background on Deep Neural Networks (DNN) topics, notably convolutional NNs (CNNs) and deep learning for object detection. Various DNN programming tools will be presented, e.g., PyTorch, Keras, Tensorflow. The hands-on programming workshop will be on PyTorch basics and target detection with Pytorch.
Part B lectures will focus on computer vision algorithms, namely on 2D target tracking, 3D target localization techniques (giving the attendants the opportunity to master state of the art video trackers), parallel GPU, multi-core CPU architectures and GPU programming (CUDA). Two programming workshops will take place. The first one will be on CUDA programming, focusing on 2D convolution algorithms. The second one will be on how to use OpenCV (the most used library for computer vision) for target tracking.
As drones execute missions (e.g., AV shooting, inspection), Part C lectures will focus on Drone mission planning and control. Before mission execution, it is best simulated, using drone mission simulation tools. Such simulations will be presented using AirSim. Additionally a programming workshop on ROS and Gazebo simulations for drones will take place.
The lectures and programming tools will provide programming skills for the various computer vision and deep learning problems encountered in drone imaging and cinematography, which is one of the main application areas of drone technologies. The same machine learning and computer vision problems do occur in other drone applications as well, e.g., for land/marine surveillance, search&rescue, building and machine inspection.
Lectures and programming workshops will be in English. PDF files will be available at the end of the course. 40 programming workstation positions will be available on a registration priority basis. Registration will stop if/when the positions are filled.
If in-depth coverage of computer vision and deep learning is desired, the audience may also want to join the short course on
‘Computer Vision and Deep Learning course’ 26-27/08/2019, Aristotle University of Thessaloniki, that is back to back to this course.
Part A (8 hours), Deep learning sample topic list
- Deep neural networks. Convolutional NNs
- Deep learning for target detection
- PyTorch basics
- Target detection with PyTorch
- Object-oriented Tensorflow in Google Colab
Part B (8 hours), Computer vision sample topic list
- 2D target tracking and 3D target localization
- Parallel GPU and multi-core CPU architectures. GPU programming
- CUDA programming
- OpenCV programming for object tracking
- Drone mission simulations
Part C (8 hours), Drone planning/control sample topic list
- Drone mission planning and control
- Gimbal control for target tracking
- Drones with ROS and Gazebo simulations
- Brain-Drone Interaction
The course will take place on 28-30 August 2019.
All lectures and workshops will take place at the School of Informatics lecture halls and labs, Aristotle University of Thessaloniki.