Zobrazeno 1 - 10
of 39
pro vyhledávání: '"Kim, Hyunjik"'
Autor:
Mirowski, Piotr, Warde-Farley, David, Rosca, Mihaela, Grimes, Matthew Koichi, Hasson, Yana, Kim, Hyunjik, Rey, Mélanie, Osindero, Simon, Ravuri, Suman, Mohamed, Shakir
Atmospheric states derived from reanalysis comprise a substantial portion of weather and climate simulation outputs. Many stakeholders -- such as researchers, policy makers, and insurers -- use this data to better understand the earth system and guid
Externí odkaz:
http://arxiv.org/abs/2407.11666
Most neural compression models are trained on large datasets of images or videos in order to generalize to unseen data. Such generalization typically requires large and expressive architectures with a high decoding complexity. Here we introduce C3, a
Externí odkaz:
http://arxiv.org/abs/2312.02753
Autor:
Mehrabian, Abbas, Anand, Ankit, Kim, Hyunjik, Sonnerat, Nicolas, Balog, Matej, Comanici, Gheorghe, Berariu, Tudor, Lee, Andrew, Ruoss, Anian, Bulanova, Anna, Toyama, Daniel, Blackwell, Sam, Paredes, Bernardino Romera, Veličković, Petar, Orseau, Laurent, Lee, Joonkyung, Naredla, Anurag Murty, Precup, Doina, Wagner, Adam Zsolt
This work studies a central extremal graph theory problem inspired by a 1975 conjecture of Erd\H{o}s, which aims to find graphs with a given size (number of nodes) that maximize the number of edges without having 3- or 4-cycles. We formulate this pro
Externí odkaz:
http://arxiv.org/abs/2311.03583
Autor:
Bauer, Matthias, Dupont, Emilien, Brock, Andy, Rosenbaum, Dan, Schwarz, Jonathan Richard, Kim, Hyunjik
Neural fields, also known as implicit neural representations, have emerged as a powerful means to represent complex signals of various modalities. Based on this Dupont et al. (2022) introduce a framework that views neural fields as data, termed *func
Externí odkaz:
http://arxiv.org/abs/2302.03130
Autor:
Kim, Hyunjik1 (AUTHOR), Ko, Dai Sik2 (AUTHOR) daisik.ko@gilhospital.com
Publikováno v:
Journal of Clinical Medicine. Oct2024, Vol. 13 Issue 19, p5908. 10p.
Autor:
Zaidi, Sheheryar, Berariu, Tudor, Kim, Hyunjik, Bornschein, Jörg, Clopath, Claudia, Teh, Yee Whye, Pascanu, Razvan
Re-initializing a neural network during training has been observed to improve generalization in recent works. Yet it is neither widely adopted in deep learning practice nor is it often used in state-of-the-art training protocols. This raises the ques
Externí odkaz:
http://arxiv.org/abs/2206.10011
Autor:
Zaidi, Sheheryar, Schaarschmidt, Michael, Martens, James, Kim, Hyunjik, Teh, Yee Whye, Sanchez-Gonzalez, Alvaro, Battaglia, Peter, Pascanu, Razvan, Godwin, Jonathan
Many important problems involving molecular property prediction from 3D structures have limited data, posing a generalization challenge for neural networks. In this paper, we describe a pre-training technique based on denoising that achieves a new st
Externí odkaz:
http://arxiv.org/abs/2206.00133
Autor:
Miao, Ning, Rainforth, Tom, Mathieu, Emile, Dubois, Yann, Teh, Yee Whye, Foster, Adam, Kim, Hyunjik
We introduce InstaAug, a method for automatically learning input-specific augmentations from data. Previous methods for learning augmentations have typically assumed independence between the original input and the transformation applied to that input
Externí odkaz:
http://arxiv.org/abs/2206.00051
It is common practice in deep learning to represent a measurement of the world on a discrete grid, e.g. a 2D grid of pixels. However, the underlying signal represented by these measurements is often continuous, e.g. the scene depicted in an image. A
Externí odkaz:
http://arxiv.org/abs/2201.12204
Subsampling is used in convolutional neural networks (CNNs) in the form of pooling or strided convolutions, to reduce the spatial dimensions of feature maps and to allow the receptive fields to grow exponentially with depth. However, it is known that
Externí odkaz:
http://arxiv.org/abs/2106.05886