Amirhossein Mashghdoust's repositories
black-and-white-pieces-Classifier-in-industry-using-PLC
I have designed a ladder PLC program that works in two modes manual and auto, and can detect white and black pieces using pre defined sensors and can separate them.
Classifying-EEG-signals-by-extracting-features-from-a-moving-time-window-using-different-ML-models
In this project I have extarcted 30 time and frequancy features from EEG signals (of left hand and right hand moving) in an espicific time window. Then using PCA i have decreased the features dimension to 10. Then I have quarried different methdos of ML: KNN(1,3,5,6), SVM(Linear kernel, Gaussian kernel), LDA, Naive bayes on different time windows.
detecting-players-in-a-football-match
In this project I have designed a program that can detect the players in a soccer match. the project includes video processing, image processing such as filtering and erosion and dilation, homographic transforms, and a simple CNN model to label the players of different teams and referee.
Kalman-Filter-Implimentation
I have implemented a kalman filter from an existing paper that can predict and estimate the behavior of a noisy dynamic system. I have used kernel trick to compute some of the equations.
creating-portrait-paintings-using-GAN
In this project I have trained a GAN moedel to generate paintings using 2000 classic paintings. the GAN can be used to genearte any 2D data.
Designing-Rectifiers-Using-Triodes
Using Simulink, I have designed a rectifier and I have shown each signal output.
detecting-arbitrary-colors-and-tracking-them-in-real-time-uses
In this project I have implemented a program that detects and tracks the center of an input color in a real-time input. One of the applications can be controling the computer's mouse using camera.
Evaluating-ML-Models-Using-Data-Generated-by-GAN
In this project we have preproccesed a dataset and generated synthetic data using GAN. Then we have compared the accuracy of different ML models on the generated data.
neuron-spike-detection-application
In this project I have implemented an application using tkinter library to classify different brain neuron types. I have used raw data in time and frequency domain (Wavelet Transform), given them to different ML models such as LDA, SVM, CNN both in 1D (raw data) and 2D (Wavelet Transform of spikes) and query each classification.
Real-Time-Road-Signs-Classsification-Using-CNN
In this project I have implemented a model which can classify the road sign photos into 6 classes. There is a video of me testing the program in the presentation power point.
modeling-and-programming-a-controller-for-an-industrial-application
In this project I have programmed a PLC program to run an industrial relay. I have also modeled the system and designed a PID controller for that.
new-approach-to-solve-the-art-gallery-problem
In this project using induction on the number of triangle faces in a simple polynomial, we tried to show that [n/3] is the optimal answer to the (pi-floodlight) art gallery problem.
note-sheet-application
In this project I have implemented a program that detects and shows the note sheet given a piano song. I use short Fourier transform to find the frequency range in each time period and then using a trick I detect all the played notes in that period.
PID-Controller-using-different-methods
I have designed different PID systems to track an arbitrary digital signal using CHR1, CHR2, ZN(ziegler nichols)