Representational Continuity for Unsupervised Continual Learning

ICLR 2022 Representational Continuity for Unsupervised Continual Learning Divyam Madaan, Jaehong Yoon, Yuanchun Li, Yunxin Liu, Sung Ju Hwang  Continual learning (CL) aims to learn a sequence of tasks without forgetting the previously acquired knowledge. However, recent CL advanc...

Set Based Stochastic Subsampling

ICML 2022 Set Based Stochastic Subsampling Bruno Andreis, Seanie Lee, A. Tuan Nguyen, Juho Lee, Eunho Yang, Sung Ju Hwang  Deep models are designed to operate on huge volumes of high dimensional data such as images. In order to reduce the volume of data these models must process,...

Sequential Reptile: Inter-Task Gradient Alignment for Multilingual Learning

ICLR 2022 Sequential Reptile: Inter-Task Gradient Alignment for Multilingual Learning Seanie Lee, Hae Beom Lee, Juho Lee, Sung Ju Hwang  Multilingual models jointly pretrained on multiple languages have achieved remarkable performance on various multilingual downstream tasks. Mor...

Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations

ICML 2022 Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations Jaehyeong Jo, Seul Lee, Sung Ju Hwang  Generating graph-structured data requires learning the underlying distribution of graphs. Yet, this is a challenging problem, and the pre...

Saliency Grafting: Innocuous Attribution-Guided Mixup with Calibrated Label Mixing

AAAI 2022 Saliency Grafting: Innocuous Attribution-Guided Mixup with Calibrated Label Mixing Joonhyung Park, June Yong Yang, Jinwoo Shin, Sung Ju Hwang, Eunho Yang  The Mixup scheme suggests mixing a pair of samples to create an augmented training sample and has gained considerab...

Rethinking the Representational Continuity: Towards Unsupervised Continual Learning

ICLR 2022 Rethinking the Representational Continuity: Towards Unsupervised Continual Learning Divyam Madaan, Jaehong Yoon, Yuanchun Li, Yunxin Liu, Sung Ju Hwang  Continual learning (CL) aims to learn a sequence of tasks without forgetting the previously acquired knowledge. ...

Rethinking the Entropy of Instance in Adversarial Training

SaTML 2022 Rethinking the Entropy of Instance in Adversarial Training Minseon Kim, Jihoon Tack, Jinwoo Shin, Sung Ju Hwang   Adversarial training, which minimizes the loss of adversarially-perturbed training examples, has been extensively studied as a solution to improv...

Real-Time Seizure Detection using EEG: A Comprehensive Comparison of Recent Approaches under a Realistic Setting

CHIL 2022 Real-Time Seizure Detection using EEG: A Comprehensive Comparison of Recent Approaches under a Realistic Setting Kwanhyung Lee, Hyewon Jeong, Seyun Kim, Donghwa Yang, Hoon-Chul Kang, Edward Choi  The Mixup scheme suggests mixing a pair of samples to create an augmented ...

Online Hyperparameter Meta-Learning with Hypergradient Distillation

ICLR 2022 Online Hyperparameter Meta-Learning with Hypergradient Distillation Hae Beom Lee, Hayeon Lee, Jaewoong Shin, Eunho Yang, Timothy Hospedales, Sung Ju Hwang    Many gradient-based meta-learning methods assume a set of parameters that do not participate in inner-optim...

Online Coreset Selection for Rehearsal-based Continual Learning

ICLR 2022 Online Coreset Selection for Rehearsal-based Continual Learning Jaehong Yoon, Divyam Madaan, Eunho Yang, Sung Ju Hwang  A dataset is a shred of crucial evidence to describe a task. However, each data point in the dataset does not have the same potential, as some of...