This is a repository fo projects of CMSC733-Classical and Deep Learning Approaches for Geometric Computer Vision in UMD. The codes are written mostly by myself with the starter code provided by the course instructors. The images in this repository belongs to the university and the course instructors.
In this project, I implemented my first CNN using TensorFlow. On the CIFAR-10 dataset, I achieved about 74% accuracy. Also, I implemented the concept of ResNet, ResNeXt and DenseNet. Along them, my ResNet achieved about 80% accuracy.