Deepmind Technologies Limited

Google Inc.

Mustafa Suleyman

Samuel Blackwell

Arun Sarath Nair

Philip Blunsom

Benjamin Kenneth Coppin

Edward Thomas Grefenstette

Olaf Ronneberger

Trevor Back

Jeffrey De Fauw

Harry Askham

Julien Robert Michel Cornebise

Xavier Glorot

Bernardino Romera-Paredes

Stanislav Nikolov

Pearse A Keane

Karl Moritz Hermann

Nenad Tomasev

Koray Kavukcuoglu

David Silver

Volodymyr Mnih

Joseph R Ledsam

Balaji Lakshminarayanan

Vedavyas Panneershelvam

Rory Fearon

Cagdas Alcicek

Alessandro De Maria

Praveen Deepak Srinivasan

Thomas Chadwick Walters

Luis Carlos Cobo Rus

Lasse Espeholt

Andrei-Alexandru Rusu

Tomas Kocisky

Marek Barwinski

Chia-Yueh Carlton Chu

Timothy Mann

Martin Szummer

Jean-Francois Crespo

William Thomas Kay

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for augmenting neural networks with an external memory. One of the methods includes providing an output derived from the neural network output for the time step as a system output for the time step; maintaining a current state of the external memory; determining, from the neural network output for the time step, memory state parameters for the time step; updating the current state of the external memory using the memory state parameters for the time step; reading data from the external memory in accordance with the updated state of the external memory; and combining the data read from the external memory with a system input for the next time step to generate the neural network input for the next time step.

Augmented recurrent neural network with external memory

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating a final classification output for an image of eye tissue. The image is provided as input to each of one or more segmentation neural networks to obtain one or more segmentation maps of the eye tissue in the image. A respective classification input is generated from each of the segmentation maps. For each of the segmentation maps, the classification input for the segmentation map is provided as input to each of one or more classification neural networks to obtain, for each segmentation map, a respective classification output from each classification neural network. A final classification output for the image is generated from the respective classification outputs for each of the segmentation maps.

Generalizable medical image analysis using segmentation and classification neural networks

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for distributed training of reinforcement learning systems. One of the methods includes receiving, by a learner, current values of the parameters of the Q network from a parameter server, wherein each learner maintains a respective learner Q network replica and a respective target Q network replica; updating, by the learner, the parameters of the learner Q network replica maintained by the learner using the current values; selecting, by the learner, an experience tuple from a respective replay memory; computing, by the learner, a gradient from the experience tuple using the learner Q network replica maintained by the learner and the target Q network replica maintained by the learner; and providing, by the learner, the computed gradient to the parameter server.

Distributed training of reinforcement learning systems

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using a machine learning model that has been trained through reinforcement learning to select a content item. One of the methods includes receiving first data characterizing a first context in which a first content item may be presented to a first user in a presentation environment; and providing the first data as input to a long-term engagement machine learning model, the model having been trained through reinforcement learning to: receive a plurality of inputs, and process each of the plurality of inputs to generate a respective engagement score for each input that represents a predicted, time-adjusted total number of selections by the respective user of future content items presented to the respective user in the presentation environment if the respective content item is presented in the respective context.

Selecting content items using reinforcement learning

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for selecting answers to questions about documents. One of the methods includes receiving a document comprising a plurality of document tokens; receiving a question associated with the document, the question comprising a plurality of question tokens; processing the document tokens and the question tokens using a reader neural network to generate a joint numeric representation of the document and the question; and selecting, from the plurality of document tokens, an answer to the question using the joint numeric representation of the document and the question.

Reading comprehension neural networks

A method for processing an image to generate a signature which is characteristic of a pattern within said image. The method comprising receiving an image; overlaying a window at multiple locations on said image to define a plurality of sub-images within said image, with each sub-image each having a plurality of pixels having a luminance level; determining a luminance value for each said sub-image, wherein said luminance value is derived from said luminance levels of said plurality of pixels; and combining said luminance values for each of said sub-images to form said signature. Said combining is such that said signature is independent of the location of each sub-image. A method of creating a database of images using said method of generating signatures is also described.

Method and apparatus for image processing

The present invention relates to a method and system for characterizing an image. The characterization may then be used to conduct a search for similar images, for example using a learning system trained using previously characterized images. A face may be identified within the image and a subsection extracted from said image which does not contain said face. At least one fixed size patch is taken from said extracted subsection; and input into said learning network to characterize said image.

Method and apparatus for characterizing an image

A method is described for processing an image to generate a signature which is characteristic of a pattern within the image. The method includes: receiving an image; overlaying a window at multiple locations on the image to define a plurality of sub-images within the image, with each sub-image each having a plurality of pixels having a luminance level; determining a luminance value for each sub-image where the luminance value is derived from the luminance levels of the plurality of pixels; and combining the luminance values for each of the sub-images to form the signature. The combining is such that the signature is independent of the location of each sub-image. A method of creating a database of images using the method of generating signatures is also described.

Growing community of inventors

London, United Kingdom

Mustafa Suleyman