Fly-Vision-Analysis-using-CNNs
Project done during Mitacs Globalink Research Internship 2018
Description
The project explores the visual acuity of fruitflies by investigating a theoretical model of its visual system.
Its visual system has been found to be very similar to a architecture of a convolutional neural network, and
hence could be compared with existing networks, such as ResNet18, ZFNet,etc.
Code Organisation
data_aug.py -- Code for augmentation of dataset of fruitfly images.
fly_eye.py -- Code for training and testing various models described in the models.py.
models.py -- Implements classes for ResNet18, fly eye model,etc. These are re-used in fly_eye.py