We use deep convolutional autoencoder to do classification and later segmentation on a subset of PASCAL VOC2009 dataset. We also apply techniques like t-SNE, UMAP, PCA for visualization of latent space.
We use 5 classes of images viz. of 'aeroplane', 'car', 'chair', 'dog' and 'bird' for our classification and segmentation task.
soutrikmachine/Classification-and-Segmentation-on-PASCAL-VOC2009
We use deep convolutional autoencoder to do classification and later segmentation on a subset of PASCAL VOC2009 dataset. We also apply techniques like t-SNE, UMAP, PCA for visualization of latent space.
Python