Signal Processing and Deep Learning Image Classification
In this project, EEG signals recorded from healthy subjects, seizure free epileptic subjects and seizure epileptic subjects were represented as image by using short time Fourier, wavelet and Hilbert-Huang transforms. Then, CNNs are trained with these images. Finally, unseen images classified with trained networks. With different configurations an accuracy of 100% is achived.
I'm an electircal engineer and PhD student. I am specialized in but not limited to: - Digital signal processing - Biomedical signal (EEG, ECG, etc) processing - Pre-processing signals for machine learning algorithm - Features extraction for machine learning algorithm - Classification with k nearest neighbor (kNN) - Classification, regression, modeling and prediction with Artificial neural network (ANN) - Classification, regression, modeling and prediction with Support vector machines (SVM) - Image/signal transformation and representation for deep learn. ing algorithms - Image to label classification with CNN - Image to image regression with CNN - Object detection with R-CNN, fast R-CNN, faster R-CNN and YOLO - Image denoising with DnCNN - Image generation with GAN and VAE and etc. - Time series prediction with LSTM - Sequence to label classification with LSTM - Sequence to sequence classification with LSTM