Self-supervised predictive coding and multimodal fusion advance patient deterioration prediction in fine-grained time resolution

ICLR 2023  (TML4H workshop)  Self-supervised predictive coding and multimodal fusion advance patient deterioration prediction in fine-grained time resolution Kwanhyung Lee, John Won, Heejung Hyun, Sangchul Hahn, Edward Choi, Joohyung Lee   Accurate time prediction of pa...

Self-Distillation for Further Pre-training of Transformers

ICLR 2023 Self-Distillation for Further Pre-training of Transformers  Seanie Lee, Minki Kang, Juho Lee, Sung Ju Hwang, Kenji Kawaguchi   Pre-training a large transformer model on a massive amount of unlabeled data and fine-tuning it on labeled datasets for diverse downs...

Scale-invariant Bayesian Neural Networks with Connectivity Tangent Kernel

ICLR 2023 Scale-invariant Bayesian Neural Networks with Connectivity Tangent Kernel  Sung-Yub Kim, Sihwan Park, Kyungsu Kim and Eunho Yang   Explaining generalizations and preventing over-confident predictions are central goals of studies on the loss landscape of n...

On the Soft-Subnetwork for Few-Shot Class Incremental Learning

ICLR 2023 On the Soft-Subnetwork for Few-Shot Class Incremental Learning Haeyong Kang, Jaehong Yoon, Sultan Rizky Hikmawan Madjid, Sung Ju Hwang, Chang D. Yoo  Inspired by Regularized Lottery Ticket Hypothesis, which states that competitive smooth (non-binary) subnetworks exist w...

Martingale Posterior Neural Processes

ICLR 2023 Martingale Posterior Neural Processes  Hyungi Lee, Eunggu Yun, Giung Nam, Edwin Fong, Juho Lee   A Neural Process (NP) estimates a stochastic process implicitly defined with neural networks given a stream of data, rather than pre-specifying priors already...

Learning Input-agnostic Manipulation Directions in StyleGAN with Text Guidance

ICLR 2023 Learning Input-agnostic Manipulation Directions in StyleGAN with Text Guidance   Yoonjeon Kim, Hyunsu Kim, Junho Kim, Yunjey Choi, Eunho Yang   With the advantages of fast inference and human-friendly flexible manipulation, image-agnostic style manip...

Exploring the Role of Mean Teacher in Self-supervised Masked Auto-Encoders

ICLR 2023 Exploring the Role of Mean Teacher in Self-supervised Masked Auto-Encoders Youngwan Lee, Jeffrey Ryan Willette, Jonghee Kim, Juho Lee, Sung Ju Hwang   Masked image modeling (MIM) has become a popular strategy for self-supervised learning (SSL) of visual representat...

Decoupled Training for Long-Tailed Classification With Stochastic Representations

ICLR 2023 Decoupled Training for Long-Tailed Classification With Stochastic Representations  Giung Nam, Sunguk Jang, Juho Lee   Decoupling representation learning and classifier learning has been shown to be effective in classification with long-tailed data. There ...

Skill-based Meta-Reinforcement Learning

ICLR 2022 Skill-based Meta-Reinforcement Learning Taewook Nam, Shao-Hua Sun, Karl Pertsch, Sung Ju Hwang, Joseph J Lim    While deep reinforcement learning methods have shown impressive results in robot learning, their sample inefficiency makes the learning of complex, long-...

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...