Deadline to register is Jun 30, 2019.
Teams can still edit your proposals during judging period.

Category: Sort By:
Regional Final
📁Machine Learning
👤Sergey Postolnik ( Robotics center "Robot.ON")
📅Oct 08, 2019
The system of automated recognition of the boundaries of territories with a cloud filtering algorithm for satellite imagery. By taking satellite images of the same terrain in different spectral channels and overlapping them with each other, determining the boundaries of certain territories and objects on it with the ability to predict parts that are covered by clouds and other interference that are between the satellites and the territory under study by using computer vision and neural networks based on FPGA. The main task is to get a vectorized map of the area with the exact contours of all the objects on it.
details »
👀 2667   💬 2
Regional Final
📁Other: Medical Care
👤Vlad-Mihai PLACINTA (Polytechnic University of Bucharest)
📅Oct 10, 2019
We propose to develop a medical FPGA-based platform to be used for human medical assessment under different living conditions and with different human subjects. The platform will be composed from two main parts: the FPGA development board and a data acquisition system which will be placed on the human’s hand as a smartwatch/smart band. Its design will embed a multiple readout: environmental humidity and temperature, EKG, heart rate, pulse oxymeter, and a custom circuit to measure the galvanic skin response. All these sensors will be controlled and monitored by using the FPGA platform and their data will be processed inside the FPGA. Advanced processing statistics and other operations like Fast Fourier Transform will be performed on the data by using the FPGA resources. The final data will be sent out via Bluetooth communication link to a LabVIEWTM graphical user interface. The entire system will be powered by a custom power supply system which includes a battery, ensuring a complete galvanic isolation of the system. Several assessment scenarios on different environment conditions will be carried out, the result along with their conclusions will be presented in this project.
details »
👀 2052   💬 1
Regional Final
📁Machine Learning
👤sumanta chaudhuri (Telecom Paristech)
📅Oct 11, 2019
Every Second Counts: Real Time Weapon Detection in Surveillance Videos

Weapon detection in videos can have a multitude of applications. Firstly as a preventive action of raising an alarm in real life use cases, next to detect violent scenes in a fictional video, as well detection of radical posts in social media. Although several research teams over the world have addressed this issue, we plan to implement this function in real-time with the Intel OpenVino Starter Kit.
details »
👀 2329   💬 2
Regional Final
📁Other: Robotics
👤Veronika Artsybasheva (National Research University Higher School of Economics)
📅Oct 07, 2019
The goal of this project is to develop a prototype of a collaborative autonomous robot guide (hereinafter referred to as the Robot) for use in educational purposes.
To achieve this goal, the development process is divided into two main parts: the server part and the robotic part. The server part serves as an interaction interface between the operator and the robotized part. On the server side, the operator creates a task that is added to the queue for execution, after which it is transferred to the robotic part for direct execution.
The Robot starts its movement from the current location to the destination, avoiding collisions with static and dynamic objects. After reaching the destination point, a sequence of actions is initiated for the tour.
details »
👀 2427   💬 7
Regional Final
📁Machine Learning
👤Pawel Czapiewski (Polish-Japanese Academy of Information Technology )
📅Oct 11, 2019
Here you have the idea of EVAA: Ecological Vertical Agriculture Assistant. The project is about building a vertical farm with the environment (i.e. lighting, water, fertilizer) controlled by AI. With the use of AI reinforcement learning and digital camera output, the system will determine optimal parameters of the environment to provide the best conditions for the plant growth. A similar approach was successfully used in October 2015 in AlphaGo to defeat World Master in Go board game [1].
details »
👀 72049   💬 5
📁Other: Robotics
👤Felix von Hundelshausen (Hundelshausen International Scool of Robotics)
📅Mar 19, 2019
to be done
details »
👀 2045   💬 1
📁Other: Measuring device
👤Volodymyr Petrushak (Khmelnitskiy national university)
📅Jul 04, 2019
The measuring device is designed to measure the dynamic viscosity of Newtonian fluids. The device will include: a mechanical system, a phase-frequency converter, an information processing device, and an information display device.
details »
👀 2281   💬 3
📁Machine Learning
👤Alexander Beasley (University of Bath)
📅Apr 13, 2019
In a world that is ever reliant on communication it is increasingly important to be able to isolate and remove sources of interference. This project will use bio inspired algorithms to rapidly identify the radiation sources and track their location. Machine learning enables the system to develop its understanding of what constitutes a radiation source using historic data.
details »
👀 1889   💬 2
📁Machine Learning
👤Frank Buss (-)
📅Jul 02, 2019
Analyzing a video from one or two cameras with low latency for a drone for obstacle avoidance, and calculating speed and direction, for controlling a drone. In combination with sensor fusing from acclerometers, gyroscopes, magnetometers and GPS data, the goal is autonomous flight path planning, flying to a destination and landing safely. Later in the drone a FPGA is small and light enough to be used in a small quadcopter, but to avoid too many fatal crashes, the OpenVINO system is perfect for developing the system, in combination with the open platform AirSim from Microsoft. The video generated from the AirSim program can be streamed over the PCI bus to the FPGA, and the FPGA can stream back the processed video with markers and debug information.
details »
👀 2357   💬 4
📁Machine Learning
👤Rafał Kozik (N/A)
📅Jun 18, 2019
The aim of my project is to investigate, if DPDK framework originally developed for rapid packet processing could also be used for real time control. As an example model reinforcement learning controller for active magnetic levitation is chosen. It will demonstrate that FPGA could extends computer IO to communicate with mechanical word and burst performance of reinforcement learning to meet real time constrains.

Active magnetic levitation is dynamic system consists of permanent magnet and electromagnetic coil. The goal is to control electromagnetic force to balance gravity and maintain magnet in set position. As this setup is unstable continuous position measurement and control adjustment is needed.

As a project platform OpenVINO Starter Kit is chosen. It provides PCI interface compatible with DPDK framework. The board will be used for two tasks. Firstly to provide in/out to the magnetic levitation sensor (optical distance sensor) and effector (electromagnetic coil). Second task will be acceleration of part of the controller computation.

Main controller will be implemented as DPDK application. Its lcore threads provide near bare metal performance that probably could be used for real time control. The board will be exposed to DPDK by raw device poll mode driver [1]. Software will provide controller and simple SCADA system.

The research part of project is to measure the latency and jitter of PCI connection to FPGA board and real-time performance of DPDK lcore.

[1] DPDK Programmer’s Guide 19. Rawdevice Library,
details »
👀 93898   💬 3
📁Digital Design
👤Ingy Alaa Hafez (Cairo University)
📅Jun 29, 2019
(i) Problem Definition:
Epilepsy is one of the most common neurological disorders, affecting millions of people worldwide, characterized by an abnormality in the brain activity which leads to recurrent seizures. So the project’s main aim is to develop a hardware chip that will be implanted in the human’s brain to predict/detect seizure and non-seizure periods.

(ii) Approach and tools/techniques:
We are going to use Machine Learning techniques and algorithms for the model development and classification between seizure and non-seizure periods, in addition to Signal Processing for feature extraction. Then the high-level language model is going to be transferred to a hardware description language model using HDL programs for hardware implementation. Finally the code will be burnt on an FPGA which is known for its sensitivity, accuracy and low power.

(iii) Overview of system modules:
1. Machine Learning algorithms for training and classification.
2. Low Level VHDL
3. FPGA for hardware testing.

(iv) Impact
We consider this project as we think it will be one of the most beneficial ideas for us and for others. It will not only help us contribute in the Artificial Intelligence field but will also help others who suffer from this brain abnormality to finally relief. We will develop a small implantable chip in the patient’s brain that will predict Epilepsy seizures so that it can be treated properly as every second matters.
Dealing with Epilepsy has a lot of impacts on the person’s body, family and education as well. For the body, epilepsy can lead to lack of body control, anxiety and depression. We will cope with epilepsy and reduce these unwanted feelings to the minimal. Diagnosis of epilepsy in children leads to stressed parents too which we will also try to eliminate. This will also help reduce the prevalence found in epileptic patients of learning disabilities and memory problems.

If this project succeeded, we think it would help millions of people suffering from epilepsy carry on with their lives normally and safely, it also won’t cost them as much as surgeries. In addition, epilepsy is not usually confined to a single area of the brain but rather it’s an epileptic network where different areas of the brain interact synchronously causing these pathological spikes or seizures. And surgery is recommended only if the seizure is confined to one area. Moreover, all other previously mentioned related devices lack accuracy, are very expensive to afford, are hard to implement and they are power hungry which means that the patient having these devices in his brain must undergo a complex and expensive surgery every two years to only change the battery of these devices. So we are extremely hopeful that our device will make a huge difference and beat all others.
details »
👀 1641   💬 2
📁Machine Learning
👤Avi Salmon (Intel)
📅Jun 28, 2019
We want toi build a real time skeleton analyser for sport analysis. The project should detect and respond in real time about human position and training advice.
details »
👀 1577   💬 2

1 2 3 4 5