Generalizing Convolutions for Deep Learning
Arguably, most excitement about deep learning revolves around the performance of convolutional neural networks and their ability to automatically extract useful features from signals. In this talk I will present work from AMLAB where we generalize these convolutions. First we study convolutions on graphs and propose a simple new method to learn embeddings of graphs which are subsequently used for semi-supervised learning and link prediction. We discuss applications to recommender systems and knowledge graphs. Second we propose a new type of convolution on regular grids based on group transformations. This generalizes normal convolutions based on translations to larger groups including the rotation group. Both methods often result in significant improvements relative to the current state of the art.
Joint work with Thomas Kipf, Rianne van den Berg and Taco Cohen.
See more on this video at https://www.microsoft.com/en-us/research/video/generalizing-convolutions-deep-learning/
Видео Generalizing Convolutions for Deep Learning канала Microsoft Research
Joint work with Thomas Kipf, Rianne van den Berg and Taco Cohen.
See more on this video at https://www.microsoft.com/en-us/research/video/generalizing-convolutions-deep-learning/
Видео Generalizing Convolutions for Deep Learning канала Microsoft Research
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
Depthwise Separable Convolution - A FASTER CONVOLUTION!Geometric Deep Learning on Graphs and Manifolds - #NIPS2017Max Welling: Generalizing Convolutions for Deep LearningGauge Equivariant Convolutional Networks and the Icosahedral CNNMichael Bronstein - Geometric deep learning on graphs: going beyond Euclidean dataGroup Equivariant CNNs beyond Roto-Translations: B-Spline CNNs on Lie Groups, Erik BekkersVariational Autoencoders"Graph Neural Networks and Applications to Deep Reinforcement Learning" Neev Parikh (Brown)Graph Convolutional Networks (GCNs) made simpleGraph databases: The best kept secret for effective AILearning SO(3) Equivariant Representations with Spherical CNNs05 Imperial's Deep learning course: Equivariance and InvarianceMIT EI Seminar - Max Welling - Learning equivariant and hybrid message passing on graphsLeveraging permutation group symmetries for designing equivariant neural networks - Haggai MaronSteve Purves - Graph Convolutional Networks for Node ClassificationYann LeCun - Graph Embedding, Content Understanding, and Self-Supervised LearningSign Language Detection using ACTION RECOGNITION with Python | LSTM Deep Learning ModelDeep Learning with Coherent Nanophotonic CircuitsJure Leskovec | Advancements in Graph Neural Networks