The following project is based on an image classification task on Kaggle.
Architecture used: Resnet-18 CNN architecture For more details on working and progress, go through: Project report: https://drive.google.com/file/d/1PYL5ILS4lxhxtE3EMT6ZQsWwBCHSq_py/view?usp=sharing
Training dataset: 450 images per class for 200 classes Test set: 10,000 images Accuracy: 70.3% Ranking: 6 out of 100 participants