|
Microsoft Research - Deep Learning Group
Fundamental research in understanding and scaling large neural networks. For example, maximal update parametrization (µP) and µTransfer, the feature learning limit of neural networks, and, more generally, the theory of Tensor Programs. Follow us:
MSR Cambridge Lecture Series: An Introduction to Graph Neural Networks ...
An Introduction to Graph Neural Networks: Models and Applications Got it now: “Graph Neural Networks (GNN) are a general class of networks that work over graphs. By representing a problem as a graph — encoding the information of individual elements as nodes and their relationships as edges — GNNs learn to capture patterns within the graph.
Project Brainwave - Microsoft Research
Project Brainwave is a deep learning platform for real-time AI inference in the cloud and on the edge. A soft Neural Processing Unit (NPU), based on a high-performance field-programmable gate array (FPGA), accelerates deep neural network (DNN) inferencing, with applications in computer vision and ...
Deep Learning: Methods and Applications - microsoft.com
Deep learning is in the intersections among the research areas of neural networks, artificial intelligence, graphical modeling, optimization, pattern recognition, and signal processing.
AnIntroductiontoNeural InformationRetrieval - microsoft.com
Neural ranking models for information retrieval (IR) use shal-low or deep neural networks to rank search results in response to a query. Traditional learning to rank models employ super-vised machine learning (ML) techniques—including neural networks—over hand-crafted IR features. By contrast, more recently proposed neural models learn representations of lan-guage from raw text that can ...
An Introduction to Neural Information Retrieval
Neural ranking models for information retrieval (IR) use shallow or deep neural networks to rank search results in response to a query. Traditional learning to rank models employ supervised machine learning (ML) techniques—including neural networks—over hand-crafted IR features. By contrast, more recently proposed neural models learn representations of language from raw text that can ...
To Bridge Neural Network Design and Real-World Performance: A Behaviour ...
The paper currently focuses on CNNs (Convolutional Neural Networks) due to their high design complexity. Particularly, this study aims to answer the following ques-tions. (i) What are the behaviour characteristics that show an inconsistent latency response to the change of OPs and memory accesses of a configuration in the design space?
Gated Graph Sequence Neural Networks - Microsoft Research
Gated Graph Sequence Neural Networks Graph-structured data appears frequently in domains including chemistry, natural language semantics, social networks, and knowledge bases. In this work, we study feature learning techniques for graph-structured inputs.
Building Neural Network Models That Can Reason
The Neural State Machine (NSM) design also emphasizes the use of a more symbolic form of internal computation, represented as attention over symbols, which have distributed representations. Such designs impose structural priors on the operation of networks and encourage certain kinds of modularity and generalization.
Neural Additive Models: Interpretable Machine Learning with Neural Nets
This hinders their applicability to high stakes decision-making domains such as healthcare. We propose Neural Additive Models (NAMs) which combine some of the expressivity of DNNs with the inherent intelligibility of generalized additive models. NAMs learn a linear combination of neural networks that each attend to a single input feature.
RenderFormer: How neural networks are reshaping 3D rendering
RenderFormer, from Microsoft Research, is the first model to show that a neural network can learn a complete graphics rendering pipeline. It’s designed to support full-featured 3D rendering using only machine learning—no traditional graphics computation required. Learn more.
Stochastic Neural Networks - Microsoft Research
In the stochastic neural network project we aim to build the next generation of deep learning models which are more data-efficient and can enable machines to learn more efficiently and eventually to be truly creative.
High-Resolution Network: A universal neural architecture for visual ...
The high-resolution network (HRNet) is a universal architecture for visual recognition. The applications of the HRNet are not limited to what we have shown above, and they are suitable to other position-sensitive vision applications, such as face alignment, face detection, super-resolution, optical flow estimation, depth estimation, and so on.
Metalearned Neural Memory: Teaching neural networks how to remember ...
In deep learning, memory is key to creating more advanced systems. Leveraging metalearning techniques, Metalearned Neural Memory uses a neural network itself to store and recall data, learning how to read from and write to memory in the process.
Neural Network Intelligence - Microsoft Research
NNI (Neural Network Intelligence) is a toolkit to help users run automated machine learning (AutoML) experiments. The tool dispatches and runs trial jobs generated by tuning algorithms to search the best neural architecture and/or hyper-parameters in different environments like local machine, remote servers and cloud.
|
