Autonomous Vehicles
Orthogonal Recurrent Neural Network on DE10-Nano (ORNN Nano)
EM014 »
In this project, we aim to train a quantized orthogonal Recurrent Neural Network (ORNN) model for speech recognition and implement its inference accelerator on DE-10 Nano using intel HLS. With an era of increased use of AI to facilitate our lives speech recognition at the edge has numerous advantages. From a tiny personal assistant in our phone to the autopilots of self-driving cars, speech recognition has shown impressive performances and has contributed to assisting our lives and making it much easier. Similarly on the other hand having a speech recognition model (which is compute and memory expensive) on an edge device like DE-10 Nano allows us to explore the potential of the Altera FPGAs and to be able to deploy such technologies (e.g., bigger speech recognition model such as conversation bots) offline with a huge power savings (thanks to the configurable architecture of FPGAs). This allows us to integrate and explore the potential usage of edge devices (like DE-10 Nano) in autopilots. Moreover, the orthogonal RNNs are the choice of our project because they have the capacity to learn long-term dependencies like LSTMs but have a much lower parameter count.
What potential expertise do I have to complete this project successfully?
I (Ussama) have 4+ years of experience in working with efficient quantized training of deep learning models i.e., MLPs, CNNs, GANs, RNNs, etc., with a research internship at Xilinx on the same topic. Currently, I am an MS/Ph.D. student at KAUST working at the intersection of Deep Learning and embedded systems i.e., “Edge AI”. Have a detailed look at my LinkedIn (https://www.linkedin.com/in/ussamazahid96/) and Hackster.io profile (https://www.hackster.io/ussamazahid96) which also features my winning project i.e., “Quad96”.
Other: Forest (Biodiversity) conservation
Wildfire Prevention - Way to Conserve Biodiversity
AP044 »
Background: From past decade, ecological imbalance is causing huge impact on biodiversity of the planet earth. To ensure sustainable future for our next generations, biodiversity conservation is an important task. According to the FAO's Global Forest Resources Assessment 2020, the world has a total forest area of 4.06 billion hectares (10.0 billion acres), which is 31% of the total land area. More than half (54%) of the world's forests is found in only five countries (Brazil, Canada, China, Russia and the USA). One among the major reasons of ecological imbalance is caused by wildfires. Human caused wildfires (Global warming, Electric line & Camp fire) and Natural wildfires (Lightning, Thunderstorms & Dead trees rubbing) are the types of wildfires. The wildfires in the above listed countries have greater impact globally. Uncontrolled wildfires cause huge threats. Fast spreading wildfires take lot of time to combat and result in huge loss of flora and fauna of the particular region and thus cause ecological imbalance. On average, more than 100,000 wildfires clear 9-10 million acres (3.6-4 million hectares) of land in the world every year. In recent years, wildfires have destroyed 12-13 million acres (4.8-5 million hectares) of land. Furthermore, there are ecological, economical & social impacts as well. Wildfires also causes rise in Global warming & respiratory disorders. There have been lot of concerns and discussions regarding this recently and the researchers are still in the process to find effective and ecofriendly, cost effective approaches to prevent and detect the occurrence of wildfire. At present there are systems that are practically implemented to detect the wildfire. These systems successfully detect forest fire but are inefficient to prevent them in advance. These systems are failing miserably to combat wildfires. Hence there is a need to raise a concern regarding the prevention of wildfire by observing the climate changes in the region and monitoring the forests. We believe our approach to this problem helps in addressing this global concern to help conserve the rich culture of wildlife.
Other: Internet of Things : Home/Industrial Automation and Security system on FPGA with hardware acceleration
AI based Smart Home Assistant and Security System
AP045 »
With the rapid emergence of IoT-based technologies in the recent era, the need for high-performance edge computing to handle data and run AI-based algorithms has increased. The objective of moving the computation from the cloud to the edge devices for compute-intensive tasks, provides increased reconfigurability, low latency, and scalability. In this project, we use Intel’s DE-10 Nano SoC FPGA, by leveraging the ARM Processor to run Linux Server, Docker containers, etc., required to host the IoT system for the premise, we offload the AI-based object detection task from the HPC to the FPGA to accelerate it. The system design uses the ARM Processor to run the Linux Server and connect to the home network hosting the Home-assistant framework, NodeRED, Grafana, and other services. On the other hand, every switch box in the house is equipped with a Wi-Fi-based Custom ESP8266 based Relay switches running Tasmota firmware which is connected to the same network. Similar edge deployments are made for getting temperature, water level in tanks, etc., All these data are communicated to the Home Assistant server running on the HPC. Using Azure Event hub integration we send the data to the Azure IoT Hub. The security system consists of CCTV Cameras which are connected to DVR and using ONVIF the stream is captured in our application for object detection and processing. We use the FPGA for Accelerating the YOLOv2 object detection algorithm. We process the results and in the event of an intruder/vehicle/motion detection, the necessary alarms/lightings/or any other actions like email/SMS/mobile alerts, etc. can be set up according to the end-user requirements. The use of FPGA in this kind of security system allows to train new data sets for uses cases like but not limited to automatic distinguished detection of house members to perform actions like opening the garage doors/preventing false-positive security alarms/ lighting based on motion detection etc., We believe that by using SoC FPGAs especially DE-10 Nano, along with Microsoft Azure IoT, we would be able to accomplish a reconfigurable, scalable and low-cost IoT setup that can be deployed in real-time.
Smart City
Solar Follower
AS028 »
A sensor based solar farm that can be installed on top of a house.
Water Related
Smart Water Conservation System
AP046 »
The project is aimed to improve water distributor's capacity so everyone can use water even during dry and drought season.
Food Related
Smart and Safe Livestock Farm Monitoring System
AP047 »
The project aims to provide real time counting and monitoring of poultry or livestock which are freed to graze on the farm's grass. Using FPGAs parallel execution, the system will be able to automate the farm operations. With image processing, the farm will also identify and alarm for any predators coming inside the farm such as snakes, lizards or wolves. The system will connect to Microsoft Azure where it sends those head count for real time information to the owner. The system will also connect to cloud services to get the weather conditions and keep the farm animals safe.
Smart City
基于FPGA的智慧教室系统
PR027 »
随着无线通信、云计算技术、人工智能技术的不断发展和成熟，物联网被广泛应用于各个产业领域，建设面向智慧校园、智慧教室的物联网云平台在高校迅速发展起来。为此我们设计并制作了一款利用Intel DE10-nano现场可编程逻辑平台的智慧教室系统。本系统充分利用了Intel DE10-nano中FPGA与ARM的高速并行运算的优势，通过在Intel DE10-nano现场可编程逻辑平台上移植linux操作系统，在其中运用人工智能技术，将摄像头对课堂的实时监测画面进行AI算法处理。处理后将识别出当前课堂学生数量，每位学生信息，并通过识别学生动作，判断出学生是否举手，是否在睡觉，是否存在在玩手机等行为，最后将所识别信息通过intel FPGA Cloud Connectivity Kit即英特尔FPGA云连接套件上传到云端，帮助教师对学生进行云端监控，实现了教师维持课堂秩序的便捷管理功能。
Industrial
UAV Intelligent Recognition System Based on Deep Convolutional Neural Network
PR028 »
随着无人机地快速发展与普及，在带来便利的同时，也出现了越来越多的“黑飞”事件，给国家和公众都带来了安全隐患和经济损失。如何反制无人机非法入侵的问题越来越受到各国的重视，而快速精确识别无人机是实现反制的关键。与通过图像识别的方法识别无人机相比，通过无人机辐射的电磁信号进行无人机的识别具有更宽泛的应用范围和更高的灵敏度。无人机所辐射的信号中，无人机与操作手之间通信的飞控信号必然存在，因此识别无人机遥控信号谱图是一个有效的解决方案。民用无人机的飞控信号一般采用跳频信号进行通信。跳扩频信号属于非平稳信号，可通过短时傅里叶变换 STFT 方法，将天线、接收机采集到的多个时刻的信号变换成信号谱图。然后利用卷积神经网络（CNN）在图像识别方面的优势，实现精准识别。
本项目基于以上分析，考虑将信号多特征提取与识别的思想适用于工程实现的无人机检测与识别算法。采用基于神经网络的无人机遥控信号识别监测算法，通过联合自适应信号检测阈值计算改进现有的阈值计算方法，并通过预处理操作来对抗窄带与宽带干扰，最后利用神经网络识别是否存在及其机型。借助 FPGA 高速处理性能，对算法进行流水线设计，完成了一套具有实际工程意义的无人机检测与识别系统，该系统具有库内多架无人机检测及型号识别，以及库外无人机检测功能。
Water Related
Sustainable Aquaphonics
AP048 »
Singapore is a highly industrialized country which is one of the top countries with high population density. Trees and vegetation can be found in every block but is limited in every multi-storey residential building like HDBs and Condominiums. The projects aims to promote local hdb/condo residents to implement and improve their existing aquariums into a sustainable aquaponics which helps to reduce the carbon footprint and reduce the water usage.
Other: Data collection and monitoring
Sustainable Agriculture:Smart Farm For Biodiversity preservation in South-Cameroon
EM016 »
The project uses a FPGA kit and sensors to collect soil pH, soil moisture, soil temperature and soil NPK in smart farm to improve agriculture data collection and analysis
Other: 智能农业
智能农业温室大棚监管系统研究
PR029 »
本设计基于物联网的三层架构：感知层、网络层、应用层，构建农业温室大棚监控管理系统，通过多种传感器实时感知农业温室大棚内部环境参数；通过WiFi将各个节点采集的数据上传到服务器；用户通过手机即可实时了解到大棚内部情況，并可进行远程设备控制。
Health
MANASIK SWASTHYA YANTRAM : SoC based mental health monitoring and recommender system
AP049 »
Mental and behavioural problems are increasing part of the health problems the world over. The burden of illness resulting from psychiatric and behavioural disorders is enormous. Although it remains grossly under-represented by conventional public health statistics, which focus on mortality rather than morbidity or dysfunction. The psychiatric disorders account for 5 of 10 leading causes of disability as measured by years lived with a disability. The overall DALYs burden for neuropsychiatric disorders increased to 15% in the year 2020. At the international level, mental health is receiving increasing importance as reflected by the WHO focus on mental health as the theme for the World Health Day (4th October 2001), World Health Assembly (15th May 2001) and the World Health Report 2001 with Mental Health as the focus.
Let’s consider Covid-19 pandemic situation, opportunities to monitor psychosocial needs and deliver support during direct patient encounters in clinical practice are greatly curtailed in this crisis by large-scale home confinement. Psychosocial services, which are increasingly delivered in primary care settings, are being offered by means of telemedicine. In the context of Covid 19, psychosocial assessment and monitoring should include queries about Covid-19 related stressors (such as exposures to infected sources, infected family members, loss of loved ones, and physical distancing), secondary adversities (economic loss, for example), psychosocial effects (such as depression, anxiety, psychosomatic preoccupations, insomnia, increased substance use, and domestic violence), and indicators of vulnerability (such as pre existing physical or psychological conditions). Some patients will need referral for formal mental health evaluation and care, while others may benefit from supportive interventions designed to promote wellness and enhance coping (such as psychoeducation or cognitive behavioral techniques). In light of the widening economic crisis and numerous uncertainties surrounding this pandemic, suicidal ideation may emerge and necessitate immediate consultation with a mental health professional or referral for possible emergency psychiatric hospitalization.
We believe that Big Problems have Simple Solutions. If someone is always monitoring you and giving you timely recommendations/suggestions it can help in a very positive way. In this project we are trying to solve this problem with a lot of positivity which is actually very important in anyone's life. Our project consists of a device connected to human body which consists of sensors which monitor different parameters from human body, gas sensor inside the living room to monitor the pollution/toxic gases in the environment, a recommender system which uses the captured images of human/video and uses ML algorithms uses Cloud and Cloud connectivity kit to monitor human behavior. Data from sensors and machine learning algorithms are put together and timely recommendation is given to the people.