Authors: Akanksha Tyagi (akanksht@usc.edu), Chinmay Naik (chinmayn@usc.edu), Kaushal Shah (kaushalk@usc.edu) We designed three deep convolutional architectures to generate multiple views from a given single view of an object in an arbitrary pose. Traditional CV techniques can handle affine transformations such as rotation, translation, scaling well. However, generating non-affine transformations like…