Zobrazeno 1 - 10
of 25
pro vyhledávání: '"Richemond, Pierre H."'
Self-predictive unsupervised learning methods such as BYOL or SimSiam have shown impressive results, and counter-intuitively, do not collapse to trivial representations. In this work, we aim at exploring the simplest possible mathematical arguments t
Externí odkaz:
http://arxiv.org/abs/2302.04817
Autor:
Bošnjak, Matko, Richemond, Pierre H., Tomasev, Nenad, Strub, Florian, Walker, Jacob C., Hill, Felix, Buesing, Lars Holger, Pascanu, Razvan, Blundell, Charles, Mitrovic, Jovana
Learning from large amounts of unsupervised data and a small amount of supervision is an important open problem in computer vision. We propose a new semi-supervised learning method, Semantic Positives via Pseudo-Labels (SemPPL), that combines labelle
Externí odkaz:
http://arxiv.org/abs/2301.05158
Autor:
Dieleman, Sander, Sartran, Laurent, Roshannai, Arman, Savinov, Nikolay, Ganin, Yaroslav, Richemond, Pierre H., Doucet, Arnaud, Strudel, Robin, Dyer, Chris, Durkan, Conor, Hawthorne, Curtis, Leblond, Rémi, Grathwohl, Will, Adler, Jonas
Diffusion models have quickly become the go-to paradigm for generative modelling of perceptual signals (such as images and sound) through iterative refinement. Their success hinges on the fact that the underlying physical phenomena are continuous. Fo
Externí odkaz:
http://arxiv.org/abs/2211.15089
Diffusion models typically operate in the standard framework of generative modelling by producing continuously-valued datapoints. To this end, they rely on a progressive Gaussian smoothing of the original data distribution, which admits an SDE interp
Externí odkaz:
http://arxiv.org/abs/2210.14784
Autor:
Chan, Stephanie C. Y., Santoro, Adam, Lampinen, Andrew K., Wang, Jane X., Singh, Aaditya, Richemond, Pierre H., McClelland, Jay, Hill, Felix
Large transformer-based models are able to perform in-context few-shot learning, without being explicitly trained for it. This observation raises the question: what aspects of the training regime lead to this emergent behavior? Here, we show that thi
Externí odkaz:
http://arxiv.org/abs/2205.05055
As humans and animals learn in the natural world, they encounter distributions of entities, situations and events that are far from uniform. Typically, a relatively small set of experiences are encountered frequently, while many important experiences
Externí odkaz:
http://arxiv.org/abs/2203.08222
Autor:
Richemond, Pierre H., Grill, Jean-Bastien, Altché, Florent, Tallec, Corentin, Strub, Florian, Brock, Andrew, Smith, Samuel, De, Soham, Pascanu, Razvan, Piot, Bilal, Valko, Michal
Bootstrap Your Own Latent (BYOL) is a self-supervised learning approach for image representation. From an augmented view of an image, BYOL trains an online network to predict a target network representation of a different augmented view of the same i
Externí odkaz:
http://arxiv.org/abs/2010.10241
Autor:
Grill, Jean-Bastien, Strub, Florian, Altché, Florent, Tallec, Corentin, Richemond, Pierre H., Buchatskaya, Elena, Doersch, Carl, Pires, Bernardo Avila, Guo, Zhaohan Daniel, Azar, Mohammad Gheshlaghi, Piot, Bilal, Kavukcuoglu, Koray, Munos, Rémi, Valko, Michal
We introduce Bootstrap Your Own Latent (BYOL), a new approach to self-supervised image representation learning. BYOL relies on two neural networks, referred to as online and target networks, that interact and learn from each other. From an augmented
Externí odkaz:
http://arxiv.org/abs/2006.07733
Deep reinforcement learning requires a heavy price in terms of sample efficiency and overparameterization in the neural networks used for function approximation. In this work, we use tensor factorization in order to learn more compact representation
Externí odkaz:
http://arxiv.org/abs/1911.11285
Autor:
Richemond, Pierre H., Guo, Yike
Recent seminal work at the intersection of deep neural networks practice and random matrix theory has linked the convergence speed and robustness of these networks with the combination of random weight initialization and nonlinear activation function
Externí odkaz:
http://arxiv.org/abs/1905.01369