Spring 2019 UnifyID AI Fellowship

Papers resulting from the fellowship

ICML 2019

1. Karttikeya Mangalam, Vinay Uday Prabhu. Do deep neural networks learn shallow learnable examples first? In Proceedings, SPML workshop, ICML-2019, June 2019, Long Beach, California, USA

2. Nicholas Roberts, Vinay Uday Prabhu, Matthew McAteer. Model weight theft with just noise inputs: The curious case of the petulant attacker. In ICML Security and Privacy of Machine Learning workshop, June 2019, Long Beach, USA. (Accepted for spotlight presentation: https://slideslive.com/38918115/model-weight-theft-with-just-noise-inputs-the-curious-case-of-the-petulant-attacker?locale=en)

3. Dian Ang Yap, Joyce Xu, Vinay Uday Prabhu, Matthew McAteer. Understanding Adversarial Robustness Through Loss Landscape Geometries. Uncertainty & Robustness in Deep Learning workshop, ICML-2019, June 2019, Long Beach, California, USA

4. Vinay Uday Prabhu, Dian Ang Yap, Alex Wang. Covering up bias with Markov blankets: A post-hoc cure for attribute prior blindness. Workshop on Invertible Neural Nets and Normalizing Flows, ICML-2019, June 2019, Long Beach, California, USA

CVPR 2019

1. Dian Ang Yap, Joyce Xu, Vinay Uday Prabhu, John Whaley. Detection of Adversarial Inputs through Entropy of Saliency Maps. CVCOPS workshop, CVPR-2019, June 2019, Long Beach, California, USA

NeurIPS 2019

1. Vinay Prabhu, Sangyun Han, John Whaley. Kannada-MNIST: Spurring the MNIST-moment for the numeral scripts in the developing world. Proceedings, ML for the Developing World (ML4D) workshop, NeurIPS-2019, Vancouver, Dec 2019, Canada

2. Dian Ang Yap, Nicholas Roberts, Vinay Uday Prabhu. Grassmannian Packings in Neural Networks: Learning with Maximal Subspace Packings for Diversity and Anti-Sparsity. Proceedings, Workshop on Information Theory and Machine Learning (ITML-2019), NeurIPS-2019, Vancouver, Dec 2019, Canada

3. Nicholas Roberts, Dian Ang Yap, Vinay Uday Prabhu. Deep Connectomics Networks: Neural network architectures inspired by neuronal networks. Proceedings, Neuro-AI Workshop, NeurIPS-2019, Vancouver, Dec 2019, Canada