V
Volodymyr Mnih
Researcher at Google
Publications - 62
Citations - 51796
Volodymyr Mnih is an academic researcher from Google. The author has contributed to research in topics: Reinforcement learning & Artificial neural network. The author has an hindex of 37, co-authored 60 publications receiving 38272 citations. Previous affiliations of Volodymyr Mnih include University of Toronto & University of Alberta.
Papers
More filters
Proceedings Article
Multiple Object Recognition with Visual Attention
TL;DR: In this paper, an attention-based model is proposed for recognizing multiple objects in images. But the model is trained with reinforcement learning to attend to the most relevant regions of the input image.
Proceedings ArticleDOI
On deep generative models with applications to recognition
TL;DR: This work uses one of the best, pixel-level, generative models of natural images–a gated MRF–as the lowest level of a deep belief network that has several hidden layers and shows that the resulting DBN is very good at coping with occlusion when predicting expression categories from face images.
Proceedings ArticleDOI
Empirical Bernstein stopping
TL;DR: This work considers problems where probabilistic guarantees are desired and demonstrates how recently-introduced empirical Bernstein bounds can be used to design stopping rules that are efficient, as well as providing upper bounds on the sample complexity of the new rules.
Proceedings Article
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
Lasse Espeholt,Hubert Soyer,Rémi Munos,Karen Simonyan,Volodymyr Mnih,Tom Ward,Yotam Doron,Vlad Firoiu,Tim Harley,Iain Dunning,Shane Legg,Koray Kavukcuoglu +11 more
TL;DR: The importance weighted actor-learner architecture (IMPALA) as discussed by the authors uses resources more efficiently in single-machine training and scales to thousands of machines without sacrificing data efficiency or resource utilisation.
Proceedings Article
Using Fast Weights to Attend to the Recent Past
TL;DR: Fast weights as discussed by the authors can be used to store temporary memories of the recent past and they provide a neurally plausible way of implementing the type of attention to the past that has recently proven helpful in sequence-to-sequence models.