Deep Learning
Encoding Robustness to Image Style via Adversarial Feature Perturbations
Adversarial training is the industry standard for producing models that are robust to small adversarial perturbations.
Myriad: a real-world testbed to bridge trajectory optimization and deep learning
GPU-compatible package unlocks potential for new algorithms to arise.
RelaySum for Decentralized Deep Learning on Heterogeneous Data
Because the workers only communicate with few neighbors without central coordination, these updates propagate progressively over the network.
Online Learning and Control of Dynamical Systems from Sensory Input
Identifying an effective model of a dynamical system from sensory data and using it for future state prediction and control is challenging.
Class-Aware Adversarial Transformers for Medical Image Segmentation
Transformers have made remarkable progress towards modeling long-range dependencies within the medical image analysis domain.