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
![](https://i.ytimg.com/vi/ICDEuqGHZ-0/default.jpg)
![Depthwise Separable Convolution - A FASTER CONVOLUTION!](https://i.ytimg.com/vi/T7o3xvJLuHk/default.jpg)
![Geometric Deep Learning on Graphs and Manifolds - #NIPS2017](https://i.ytimg.com/vi/LvmjbXZyoP0/default.jpg)
![Max Welling: Generalizing Convolutions for Deep Learning](https://i.ytimg.com/vi/TOfg-JlLILA/default.jpg)
![Gauge Equivariant Convolutional Networks and the Icosahedral CNN](https://i.ytimg.com/vi/wZWn7Hm8osA/default.jpg)
![Michael Bronstein - Geometric deep learning on graphs: going beyond Euclidean data](https://i.ytimg.com/vi/b187J4ndZWY/default.jpg)
![Group Equivariant CNNs beyond Roto-Translations: B-Spline CNNs on Lie Groups, Erik Bekkers](https://i.ytimg.com/vi/rakcnrgX4oo/default.jpg)
![Variational Autoencoders](https://i.ytimg.com/vi/9zKuYvjFFS8/default.jpg)
!["Graph Neural Networks and Applications to Deep Reinforcement Learning" Neev Parikh (Brown)](https://i.ytimg.com/vi/0xc_HldEL-s/default.jpg)
![Graph Convolutional Networks (GCNs) made simple](https://i.ytimg.com/vi/2KRAOZIULzw/default.jpg)
![Graph databases: The best kept secret for effective AI](https://i.ytimg.com/vi/2ZzGMzitNgo/default.jpg)
![Learning SO(3) Equivariant Representations with Spherical CNNs](https://i.ytimg.com/vi/Y86rzE4UzKs/default.jpg)
![05 Imperial's Deep learning course: Equivariance and Invariance](https://i.ytimg.com/vi/a4Quhf9NhMY/default.jpg)
![MIT EI Seminar - Max Welling - Learning equivariant and hybrid message passing on graphs](https://i.ytimg.com/vi/hUrbS1BhBWc/default.jpg)
![Leveraging permutation group symmetries for designing equivariant neural networks - Haggai Maron](https://i.ytimg.com/vi/k3E3XyMtQcA/default.jpg)
![Steve Purves - Graph Convolutional Networks for Node Classification](https://i.ytimg.com/vi/LFSR27BaNIQ/default.jpg)
![Yann LeCun - Graph Embedding, Content Understanding, and Self-Supervised Learning](https://i.ytimg.com/vi/UGPT64wo7lU/default.jpg)
![Sign Language Detection using ACTION RECOGNITION with Python | LSTM Deep Learning Model](https://i.ytimg.com/vi/doDUihpj6ro/default.jpg)
![Deep Learning with Coherent Nanophotonic Circuits](https://i.ytimg.com/vi/Zu_2zz41Yno/default.jpg)
![Jure Leskovec | Advancements in Graph Neural Networks](https://i.ytimg.com/vi/YhKUgh0XY50/default.jpg)