📁Other: Combination of I.O.T , Embedded system and Man to Machine communication
(Gujarat Technological University)
📅Apr 28, 2018
Using sixth sense technology an ordinary person can enhance their communication with outer world. Our idea is to create a device which is helpful for those unfortunate people who are unable to speak/talk, by providing them an easy way to communicate with outer world by using their hand gestures with help of sixth sense technology. Because of affordable price and less number of components at the user side, this device will provide an efficient way of communication. By taking the existing advantages of sixth sense technology like mobile excess and faster data exchange, we will use the FPGA board as a memory device and also as a medium between user and a server and as an image processor . The server has all the data about gestures and respective speech synthesize and will communicate with FPGA board at real time basis using IOT technology. If one does not have no internet connection, then there is no need to worry!! The FPGA board will have a storage which also has all the speech and gestures data. So ultimately this device will work efficiently in both mode, Online and Offline!
📁Internet of Things
👤Dr. Bijoy Kumar Upadhyaya
(Tripura Institute of technology)
📅Jan 30, 2018
The project is focused on the security of a citizen whether inside or outside of a house. The project is enhanced with Bluetooth camera, GPS, Audio recording module making it exclusive to ensure the security of an individual by alerting the concerned authority in smart way.
📁High Performance Computing
📅May 02, 2018
Adders are most important part in a processor working as they are present in ALU unit and perform all arithmetic and logical operations . The faster the adders involving the faster the results we get , As we are indeed trying to get faster and high end results in today technology . we are need of the fastest adders to make faster computations.
In this project , we are using a Brent kung adder and Binary excess converter( B E C) in a carry select adder instead of the regular use of ripple carry adders it is the Existing method . By replacing ripple carry adders and placing Brent kung adder along with binary excess converter , we tend to get faster results. As the carry propagation and generation time has reduced from ripple carry adder to Brent kung adder as it is parallel prefix adder works like a tree based and Binary excess converter on other hand will add 1 to the result making the faster computation . So with the use of Brent kung adder and Binary excess converter we get faster computational environment. In future we are intending to use more faster approach of adders involving use of high end addition techniques like vedic maths or use of other faster computational methods. Finally , Our method not only makes computation faster but also consumes less power as it uses less number of logics than existing results in lower power consumption and these can be implemented in FPGA.
📁Other: Orientation Estimation
(Universitas Gadjah Mada)
📅May 10, 2018
We proposed an Attitude and Heading Reference Systems (AHRS) coprocessor from
tri-axis Magnetic, Angular Rate, and Gravity (MARG) and Inertial measurement Unit (IMU) sensor using Madgwick’s
AHRS sensor fusion algorithm. By relieving processor-intensive tasks from the
primary processor, coprocessors can accelerate overall system performance.
📁Other: COMBINATION OF I.O.T , HPC ,MACHINE LEARNING
(GUJARAT TECHNOLOGICAL UNIVERSITY)
📅May 04, 2018
THE PROBLEM FACED BY MANY OF THE PEOPLE IN POSITIONING/MOVING HEAVY EQUIPMENT IN DIFFERENT TERRAINS (ESPECIALLY STAIRS AS NOTICED), SO USING MECH. DESIGN, IOT AND MACHINE LEARNING ONE SOLUTION IS TRIED.
📁Other: Deep Learning,Image Processing
(National Institute of Technology Tiruchirapalli)
📅Dec 15, 2017
Recognition of emotion can be done through different modalities, such as speech, facial expression, body gesture etc.Emotion recognition through facial expression has been an interesting area in the last few decades.Facial expressions recognition(FER) play an integral part in conveying our thoughts. FER systems are very useful in human-computer interaction, human behaviour analysis, helping autistic children etc. Because FER systems cannot capture the face in the same orientation, size etc. all the time an unconstrained facial expression recognition system is developed on FPGA.The objective of the project work is to design a less complex FER systems using binary auto-encoders(BAE) and binarized neural networks(BNN) as the classifier supporting real time performance without compromising the accuracy.
(Jawaharlal Technological University Ananthapur)
📅Dec 19, 2017
An all-digital ON-chip process sensor using a ratioed inverter-based ring oscillator is proposed. Two types of the ratioed inverter-based ring oscillators, nMOS and pMOS types, are proposed to sense process variation. The
nMOS (pMOS)-type ring oscillator is designed to improve its sensitivity to the process variation in the nMOS (pMOS) transistors using the ratioed inverter that consists of only nMOS (pMOS) transistors. A compact process sensor can be realized using only these two types of ring oscillators. For a suitable ON-chip implementation, the output of the proposed process sensor is provided with a digital code. The proposed process sensor is fabricated using a 0.13-μm CMOS technology. Measurement results from 30 fabricated chips show that all chips have the same process corner. To verify whether the proposed sensor can properly sense all the process corners, the threshold voltage of the fabricated chips is shifted by body biasing. The verification results show that the measured code error compared with the post layout simulation is less than 2.92%.
📁High Performance Computing
📅May 07, 2018
We plan to use the FPGA to perform real time identification and localisation of moving objects in the field of view of a SONAR based sensor platform mounted on a vehicle to alert drivers of near-by wildlife.
In active mode, the SONAR system will use a single speaker (transmitter) to generate a series of specific sound waves.
Reflections from these sound waves will then be recorded using an array of 64 individual microphones (receivers).
The sampled data from the microphones will be processed through a combination of time-delay beamforming and correlation with the transmitted waveform.
The result of the correlated waveforms will also be fed through a Fast Fourier Transform to provide a Doppler result, allowing detection of any moving objects and their relative velocities.
These will be fed to a Neural Network trained to identify wildlife responses.
Configuration and Communication with the SONAR platform will be performed through use of the HPS and the Ethernet port, allowing the sensor to forward warnings to the driver of the vehicle.
(Velammal College of Engineering and Technology)
📅May 11, 2018
The jasminum flower is classified depending on their quality by extracting colour, shape, texture of the flower using various techniques and performing tensor flow algorithm for machine learning process.
1. Currently, flowers are classified manually based on their quality, which is a time consuming process so that the flowers lose their freshness.
2. Accuracy of classification is poor.
The following are the existing difficulties, which could be altered by automation proposed in this project.
The system should be designed in such a way that there must a sample collector where the jasminum flower is collected directly from the farm. The flower is allowed to pass on a conveyor belt. When it enters the image capturing chamber, the image of the flower is captured and given to the microcontroller of raspberry pi then the image is subjected for feature extraction. The colour, shape, texture of the flower could be identified using the following algorithms: Colour and Edge Diversity Descriptor (CEDD), Hue Saturation Value (HSV) colour space, Linear Binary Pattern (LBP), Zernike moments. The extracted features are fed as inputs to the tensor flow. The tensor flow is open source software by Google where the machine learning and deep learning is performed. The concept of machine learning became easy after the occurrence of tensor flow which has in-build support for deep learning, tools to assemble neural networks, mathematical functions for neural networks. Once the output is generated, it is received by the microcontroller and opens the switch for the following category via switch controller. The flower that is moving on a conveyor belt falls into the box of its category since the switch was open. Thus, the flowers could be classified accordingly. The light intensity controller is used to control the intensity or brightness while capturing the image in the image capturing chamber. The multiplexer switch and the light intensity controller could be programmed in FPGA board DE10 which could be efficient.
Thus, the jasminum flower is segregated according to its quality with great accuracy using this methodology. Machine learning made the life of people easier and comfortable.
(Velammal College of Engineering and Technology)
📅Apr 29, 2018
Today the world is moving towards a digital cashless economy. Almost every country is promoting digital transactions to bring every cash-flow transparent and curb black money and underground black market. Whenever we speak about digital transaction it is either through net banking, debit cards or through online wallet payments. All the above three methods require an internet connection to carry out. Consider a situation where we need to pay for a vegetable seller along with a roadside stall where neither he can't afford a smartphone nor the internet access or in a situation where we are in a remote place in somewhere deep into the woods and we need to send some money to someone nearby. In both cases, the digital transaction is almost impossible and we need some cold cash in our wallet.
We cannot move to a stable digital economy unless we find the solution to these problems. So, to solve these issues and to move to an even more advanced economy we propose the Electronic Digital Wallet. As the name Implies this is an electronic wallet in which we can store our money in a regular wallet but in electronic form. This wallet will be compatible, will fit right into our pocket and can be taken with us everywhere. The specialty of this wallet is that we don’t need any internet access to make any payments, so it can easily replace our regular wallet.
Every citizen in this world can get benefited by this if they have a valid bank account. Governments in every country can implement this concept on the large scale which will promote to a more green and digital economy. The only thing is that this technology should be implemented in every ATM, and a bank should provide this wallet to their customers.
Outline of Working:
First of all, it will be powered by a battery. So, we need to turn it ON only when we need to make any payment. Otherwise it will be in OFF state which in turn makes battery to run for a very long time.
There are four modes of operation Send, Receive, Load, Transfer which will be explained in detail later in the consequent section. Users can do any of the following options by selecting the appropriate button in the menu.
📁High Performance Computing
(Motilal Nehru National Institute of Technology Allahabad)
📅Apr 30, 2018
This project is aimed at developing a low cost print circuit board (PCB) Frequency Modulated Continuous Wave (FMCW) radar system.
An important advantage of radars over camera and light-detection-and-ranging (LIDAR)-based systems is that radars are relatively immune to environmental conditions (such as the effects of rain, dust, and smoke). Because FMCW radars transmit a specific signal (called a chirp) and
process the reflections, they can work in complete darkness and also bright daylight (radars are not affected by glare). When compared with ultrasound, radars typically have a much longer range and much faster time of transit for their signals. This makes this sensor very useful in applications like search and rescue, mapping, navigation and automotive. Most of the components are found in small footprints which can be easily mounted on a custom PCB.
Simple un-modulated continuous wave radar is only able to detect speed of any object moving in front of the transceiver using Doppler effect. To detect range (distance) it is modulated to produce a chirp signal using VCO and a function generator. Now because of time difference between the produced signal and reflected signal there exist a frequency component out from the mixer which is the difference between the 2 signals. If there are multiple objects in front of the radar the output from the mixer is sum of different frequency components which can be obtained through Fourier transform of the output of mixer. It requires an RF/microwave transceiver circuit and ADC front-end circuit to obtain the data for further digital processing.
If an application processor is used for Fourier transforms (FFT), it takes more time than dedicated hardware processing unit severely effecting the frame rate . In this case the hardware processing unit would be the FPGA. The FPGA would provide the final frequency components to the CPU using dedicated high speed intra-chip communication. This will off-load the CPU to process other data like images from camera, computing ranges, solving non-linear equations.
A lot of the implementations have a rotating radar which construct a 2-D field in azimuth plane using accumulation of 1-D points.
For this we would experiment a static radar using concept of two separate radar sensors (Array of 2 transceivers) providing data to a same processor. This would allow to localize object in azimuth plane with a static radar. This has advantage of having a small hardware, fast scanning rate and more reliability.
👤Chin Jian Qee
(University Tunku Abdul Rahman)
📅Jan 30, 2018
Machine Learning (ML) is going to change the way on how the world solve problems. Conventionally, before the rise of Artificial Intelligence (AI), almost all problems were solved by using equations. AI serves to break that barrier and solve problems without the use of equation. Instead, it solve problems based on techniques such as naive Bayes, logistic equations, neural nets and more. In other words, machine learning learns from a huge chunk of input data to generate appropriate set of parameters for a given ML model. Thereafter, it will produce a reasonable output data based on the model that it had learned.
This project proposes a big data analytics approach with the use of Artificial Neural Network (ANN) implemented in FPGA. The purpose of implementing this in FPGA is because FPGA is known to be reconfigurable in terms of hardware. Therefore, each and every neurons can be configured in a way such that it can achieve the best performance (speed and accuracy) possible. Besides, parallel processing can also be done in FPGA so that the processing of each neurons can be done simultaneously, unlike the software approach. The parameters for this neurons (number of neurons, number of layers, activation function of each layer) can also be configured by users to adjust to their applications.