%0 Journal Article %1 dutordoir2019translation %A Dutordoir, Vincent %A van der Wilk, Mark %A Artemev, Artem %A Tomczak, Marcin %A Hensman, James %D 2019 %K gaussian-proceses geometry readings %T Translation Insensitivity for Deep Convolutional Gaussian Processes %U http://arxiv.org/abs/1902.05888 %X Deep learning has been at the foundation of large improvements in image classification. To improve the robustness of predictions, Bayesian approximations have been used to learn parameters in deep neural networks. We follow an alternative approach, by using Gaussian processes as building blocks for Bayesian deep learning models, which has recently become viable due to advances in inference for convolutional and deep structure. We investigate deep convolutional Gaussian processes, and identify a problem that holds back current performance. To remedy the issue, we introduce a translation insensitive convolutional kernel, which removes the restriction of requiring identical outputs for identical patch inputs. We show empirically that this convolutional kernel improves performances in both shallow and deep models. On MNIST, FASHION-MNIST and CIFAR-10 we improve previous GP models in terms of accuracy, with the addition of having more calibrated predictive probabilities than simple DNN models.

@article{dutordoir2019translation, abstract = {Deep learning has been at the foundation of large improvements in image classification. To improve the robustness of predictions, Bayesian approximations have been used to learn parameters in deep neural networks. We follow an alternative approach, by using Gaussian processes as building blocks for Bayesian deep learning models, which has recently become viable due to advances in inference for convolutional and deep structure. We investigate deep convolutional Gaussian processes, and identify a problem that holds back current performance. To remedy the issue, we introduce a translation insensitive convolutional kernel, which removes the restriction of requiring identical outputs for identical patch inputs. We show empirically that this convolutional kernel improves performances in both shallow and deep models. On MNIST, FASHION-MNIST and CIFAR-10 we improve previous GP models in terms of accuracy, with the addition of having more calibrated predictive probabilities than simple DNN models.}, added-at = {2020-01-26T03:47:53.000+0100}, author = {Dutordoir, Vincent and van der Wilk, Mark and Artemev, Artem and Tomczak, Marcin and Hensman, James}, biburl = {https://www.bibsonomy.org/bibtex/2fa1f1e376330196b2cb7674da5a84adb/kirk86}, description = {[1902.05888] Translation Insensitivity for Deep Convolutional Gaussian Processes}, interhash = {49ed190e4f100500b1420fc2495ebf03}, intrahash = {fa1f1e376330196b2cb7674da5a84adb}, keywords = {gaussian-proceses geometry readings}, note = {cite arxiv:1902.05888}, timestamp = {2020-01-26T03:47:53.000+0100}, title = {Translation Insensitivity for Deep Convolutional Gaussian Processes}, url = {http://arxiv.org/abs/1902.05888}, year = 2019 }