TY - GEN
T1 - Image Decomposition and Classification Through a Generative Model
AU - Yao, Houpu
AU - Regan, Malcolm
AU - Yang, Yezhou
AU - Ren, Yi
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/9
Y1 - 2019/9
N2 - We demonstrate in this paper that a generative model can be designed to perform classification tasks under challenging settings, including adversarial attacks and input distribution shifts. Specifically, we propose a conditional variational autoencoder that learns both the decomposition of inputs and the distributions of the resulting components. During test, we jointly optimize the latent variables of the generator and the relaxed component labels to find the best match between the given input and the output of the generator. The model demonstrates promising performance at recognizing overlapping components from the multiMNIST dataset, and novel component combinations from a traffic sign dataset. Experiments also show that the proposed model achieves high robustness on MNIST and NORB datasets, in particular for high-strength gradient attacks and non-gradient attacks.
AB - We demonstrate in this paper that a generative model can be designed to perform classification tasks under challenging settings, including adversarial attacks and input distribution shifts. Specifically, we propose a conditional variational autoencoder that learns both the decomposition of inputs and the distributions of the resulting components. During test, we jointly optimize the latent variables of the generator and the relaxed component labels to find the best match between the given input and the output of the generator. The model demonstrates promising performance at recognizing overlapping components from the multiMNIST dataset, and novel component combinations from a traffic sign dataset. Experiments also show that the proposed model achieves high robustness on MNIST and NORB datasets, in particular for high-strength gradient attacks and non-gradient attacks.
KW - Generative model
KW - adversarial defense
KW - classification
UR - http://www.scopus.com/inward/record.url?scp=85076814667&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85076814667&partnerID=8YFLogxK
U2 - 10.1109/ICIP.2019.8802991
DO - 10.1109/ICIP.2019.8802991
M3 - Conference contribution
AN - SCOPUS:85076814667
T3 - Proceedings - International Conference on Image Processing, ICIP
SP - 400
EP - 404
BT - 2019 IEEE International Conference on Image Processing, ICIP 2019 - Proceedings
PB - IEEE Computer Society
T2 - 26th IEEE International Conference on Image Processing, ICIP 2019
Y2 - 22 September 2019 through 25 September 2019
ER -