Implemented a multi-label image classifier on the PASCAL VOC 2007 dataset. Designed and trained a deep convolutional network to predict a binary present/absent image-level label for each of the 20 PASCAL classes using PyTorch
Data Setup (Local)
Once you have downloaded the zip file, execute the download_data script provided: ./download_data.sh
The classifier.py contains the Convolutional Neural Network used for prediction of Binary Lables for each of the 20 PASCAL classes
The MP2_P1.ipynb contains the step-by-step preprocessing and execution of the classifier on the data. It reaches a max mAP of 0.45