no code implementations • 18 Jan 2024 • Kichang Lee, Songkuk Kim, JeongGil Ko
Federated learning are inherently hampered by data heterogeneity: non-iid distributed training data over local clients.
no code implementations • 5 Jun 2023 • Hanju Yoo, Linglong Dai, Songkuk Kim, Chan-Byoung Chae
Semantic communications have shown promising advancements by optimizing source and channel coding jointly.
no code implementations • 11 Oct 2022 • Juyeop Kim, Junha Park, Songkuk Kim, Jong-Seok Lee
In this paper, we focus on the phenomenon that Transformers show higher robustness against corruptions than CNNs, while not being overconfident.
no code implementations • 8 May 2022 • Hanju Yoo, Taehun Jung, Linglong Dai, Songkuk Kim, Chan-Byoung Chae
Semantic communications are expected to enable the more effective delivery of meaning rather than a precise transfer of symbols.
3 code implementations • ICLR 2022 • Namuk Park, Songkuk Kim
In particular, we demonstrate the following properties of MSAs and Vision Transformers (ViTs): (1) MSAs improve not only accuracy but also generalization by flattening the loss landscapes.
2 code implementations • 26 May 2021 • Namuk Park, Songkuk Kim
Neural network ensembles, such as Bayesian neural networks (BNNs), have shown success in the areas of uncertainty estimation and robustness.
no code implementations • 25 Sep 2019 • Namuk Park, Taekyu Lee, Songkuk Kim
Instead of generating separate prediction for each data sample independently, this model estimates the increments of prediction for a new data sample from the previous predictions.
1 code implementation • 12 Jul 2019 • Namuk Park, Taekyu Lee, Songkuk Kim
The computational cost of this model is almost the same as that of non-Bayesian NNs.
no code implementations • DSTC Workshop 2017 • Chanyoung Park, Kyungduk Kim, Songkuk Kim
Dialog Breakdown Detection Challenge 3 of Dialog System Technology Challenge 6