In Project 1 I display if a head turns left or right using OpenCV algorithm that detects these action from a set of video frames. Here I train the algorithm using pre-recorded video files. This pipeline can therefore be used to build a application that is trained on videos from multiple sources.
In Project 2 I predict the colourisation of histology samples from their grey-scale images using a conditional GAN.
In Project 3 Here I implement a LRP for a CNN that classifies the type of brain tumor. The LRP operates by propagating the prediction backward in the neural network and is a function of neuron activations and weights as explained in https://link.springer.com/chapter/10.1007/978-3-030-28954-6_10.
In Project 4 This project explores the correlation between exercise habits and heart rate health using data collected from smartwatches. Using LSTM (Long Short-Term Memory) models, it aims to understand how variations in exercise time and energy expenditure impact heart rate. By investigating this relationship, we can uncover insights into whether regular exercise contributes to maintaining a healthy heart rate.