PT Unknown
AU Chenshen Wu
   Luis Herranz
   Xialei Liu
   Joost Van de Weijer
   Bogdan Raducanu
TI Memory Replay GANs: Learning to Generate New Categories without Forgetting
BT 32nd Annual Conference on Neural Information Processing Systems
PY 2018
BP 5966
EP 5976
AB Previous works on sequential learning address the problem of forgetting in discriminative models. In this paper we consider the case of generative models. In particular, we investigate generative adversarial networks (GANs) in the task of learning new categories in a sequential fashion. We first show that sequential fine tuning renders the network unable to properly generate images from previous categories (ie forgetting). Addressing this problem, we propose Memory Replay GANs (MeRGANs), a conditional GAN framework that integrates a memory replay generator. We study two methods to prevent forgetting by leveraging these replays, namely joint training with replay and replay alignment. Qualitative and quantitative experimental results in MNIST, SVHN and LSUN datasets show that our memory replay approach can generate competitive images while significantly mitigating the forgetting of previous categories.
ER