Publications (22)
Robust Topological Orderings for Directed Graphs
James Smith
We modify the Pearce-Kelly algorithm that maintains a topological ordering for a directed acyclic graph in order to allow cycles to be tolerated. Cycles make topological orderings…
Memory-Efficient Semi-Supervised Continual Learning: The World is its Own Replay Buffer
James Smith, Jonathan Balloch, Yen-Chang Hsu +1
Rehearsal is a critical component for class-incremental continual learning, yet it requires a substantial memory budget. Our work investigates whether we can significantly reduce t…
Dynamic Epsilon Scheduling: A Multi-Factor Adaptive Perturbation Budget for Adversarial Training
Alan Mitkiy, James Smith, Myungseo wong +3
Adversarial training is among the most effective strategies for defending deep neural networks against adversarial examples. A key limitation of existing adversarial training appro…
State/event based versus purely Action or State based Logics
James Smith
Although less studied than purely action or state based logics, state/event based logics are becoming increasingly important. Some systems are best studied using structures with in…
Proving Properties of Rich Internet Applications
James Smith
We introduce application layer specifications, which allow us to reason about the state and transactions of rich Internet applications. We define variants of the state/event based…
Unsupervised Progressive Learning and the STAM Architecture
James Smith, Cameron Taylor, Seth Baer +1
We first pose the Unsupervised Progressive Learning (UPL) problem: an online representation learning problem in which the learner observes a non-stationary and unlabeled data strea…
Lifelong Wandering: A realistic few-shot online continual learning setting
Mayank Lunayach, James Smith, Zsolt Kira
Online few-shot learning describes a setting where models are trained and evaluated on a stream of data while learning emerging classes. While prior work in this setting has achiev…
Operational Concurrency Control in the Face of Arbitrary Scale and Latency
James Smith
We present for the first time a complete solution to the problem of proving the correctness of a concurrency control algorithm for collaborative text editors against the standard c…
Pulsed laser deposition of single phase n- and p-type Cu2O thin films with low resistivity
Syed Farid Uddin Farhad, David Cherns, James Smith +2
Low resistivity (~3-24 mOhm.cm) with tunable n- and p-type phase pure Cu2O thin films have been grown by pulsed laser deposition at 25-200 0C by varying the background oxygen parti…
Turtles, Hats and Spectres: Aperiodic structures on a Rhombic tiling
James Smith
These notes derive aperiodic monotiles (arXiv:2303.10798) from a set of rhombuses with matching rules. This dual construction is used to simplify the proof of aperiodicity by consi…
On the Adversarial Robustness of Quantized Neural Networks
Micah Gorsline, James Smith, Cory Merkel
Reducing the size of neural network models is a critical step in moving AI from a cloud-centric to an edge-centric (i.e. on-device) compute paradigm. This shift from cloud to edge…
Closed ideals of operators on the Baernstein and Schreier spaces
Niels Jakob Laustsen, James Smith
We study the lattice of closed ideals of bounded operators on two families of Banach spaces: the Baernstein spaces for and the Schreier spaces for $1\le p<…
The Jasper Framework: Towards a Platform Independent, Formal Treatment of Web Programming
James Smith
This paper introduces Jasper, a web programming framework which allows web applications to be developed in an essentially platform indepedent manner and which is also suited to a f…
Global time estimates for solutions to equations of dissipative type
Michael Ruzhansky, James Smith
Global time estimates of Lp-Lq norms of solutions to general strictly hyperbolic partial differential equations are considered. The case of special interest in this paper are equat…
Eliminating Left Recursion without the Epsilon
James Smith
The standard algorithm to eliminate indirect left recursion takes a preventative approach, rewriting a grammar's rules so that indirect left recursion is no longer possible, rather…
Gender Prediction in Social Media
James Smith
In this paper, we explore the task of gender classification using limited network data with an application to Fotolog. We take a heuristic approach to automating gender inference b…
A Closer Look at Knowledge Distillation with Features, Logits, and Gradients
Yen-Chang Hsu, James Smith, Yilin Shen +2
Knowledge distillation (KD) is a substantial strategy for transferring learned knowledge from one neural network model to another. A vast number of methods have been developed for…
Finer-grained Locking in Concurrent Dynamic Planar Convex Hulls
K. Alex Mills, James Smith
The convex hull of a planar point set is the smallest convex polygon containing each point in the set. The dynamic convex hull problem concerns efficiently maintaining the convex h…
Secret Sharing for Generic Theoretic Cryptography
James Smith
Sharing a secret efficiently amongst a group of participants is not easy since there is always an adversary / eavesdropper trying to retrieve the secret. In secret sharing schemes,…
Dispersive and Strichartz estimates for hyperbolic equations with constant coefficients
Michael Ruzhansky, James Smith
Dispersive and Strichartz estimates for solutions to general strictly hyperbolic partial differential equations with constant coefficients are considered. The global time decay est…
System Design for an Integrated Lifelong Reinforcement Learning Agent for Real-Time Strategy Games
Indranil Sur, Zachary Daniels, Abrar Rahman +16
As Artificial and Robotic Systems are increasingly deployed and relied upon for real-world applications, it is important that they exhibit the ability to continually learn and adap…
Always Be Dreaming: A New Approach for Data-Free Class-Incremental Learning
James Smith, Yen-Chang Hsu, Jonathan Balloch +3
Modern computer vision applications suffer from catastrophic forgetting when incrementally learning new concepts over time. The most successful approaches to alleviate this forgett…