Search
Results
-
Adversarial Training for Free!
(2019)Advances in Neural Information Processing Systems 32Adversarial training, in which a network is trained on adversarial examples, is one of the few defenses against adversarial attacks that withstands strong attacks. Unfortunately, the high cost of generating strong adversarial examples makes standard adversarial training impractical on large-scale problems like ImageNet. We present an algorithm that eliminates the overhead cost of generating adversarial examples by recycling the gradient ...Conference Paper -
Active Learning for Student Affect Detection
(2019)Proceedings of the 12th International Conference on Educational Data Mining, EDM 2019, Montréal, Canada, July 2-5, 2019. International Educational Data Mining Society (IEDMS) 2019“Sensor-free” detectors of student affect that use only student activity data and no physical or physiological sensors are cost-effective and have potential to be applied at large scale in real classrooms. These detectors are trained using student affect labels collected from human observers as they observe students learn within intelligent tutoring systems (ITSs) in real classrooms. Due to the inherent diversity of student activity and ...Conference Paper -
Visualizing the Loss Landscape of Neural Nets
(2018)NIPS'18: Proceedings of the 32nd International Conference on Neural Information Processing SystemsNeural network training relies on our ability to find "good" minimizers of highly non-convex loss functions. It is well-known that certain network architecture designs (e.g., skip connections) produce loss functions that train easier, and well-chosen training parameters (batch size, learning rate, optimizer) produce minimizers that generalize better. However, the reasons for these differences, and their effect on the underlying loss ...Conference Paper -
An Out-of-Sample Extension for Wireless Multipoint Channel Charting
(2019)Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering ~ Cognitive Radio-Oriented Wireless Networks 14th EAI International Conference, CrownCom 2019, Poznan, Poland, June 11–12, 2019, ProceedingsChannel-charting (CC) is a machine learning technique for learning a multi-cell radio map, which can be used for cognitive radio-resource-management (RRM) problems. Each base-station (BS) extracts features from the channel-state-information samples (CSI) from transmissions of user-equipment (UE) at different unknown locations. The multi-path channel components are estimated and used to construct a dissimilarity matrix between CSI samples ...Conference Paper -
Deep Unfolding for Communications Systems: A Survey and Some New Directions
(2020)2019 IEEE International Workshop on Signal Processing Systems (SiPS)Deep unfolding is a method of growing popularity that fuses iterative optimization algorithms with tools from neural networks to efficiently solve a range of tasks in machine learning, signal and image processing, and communication systems. This survey summarizes the principle of deep unfolding and discusses its recent use for communication systems with focus on detection and precoding in multi-antenna (MIMO) wireless systems and belief ...Conference Paper -
Biconvex Relaxation for Semidefinite Programming in Computer Vision
(2016)Lecture Notes in Computer Science ~ Computer Vision – ECCV 2016 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part VISemidefinite programming (SDP) is an indispensable tool in computer vision, but general-purpose solvers for SDPs are often too slow and memory intensive for large-scale problems. Our framework, referred to as biconvex relaxation (BCR), transforms an SDP consisting of PSD constraint matrices into a specific biconvex optimization problem, which can then be approximately solved in the original, low-dimensional variable space at low complexity. ...Conference Paper