Dynamic Filter Networks

@article{Brabandere2016DynamicFN,
  title={Dynamic Filter Networks},
  author={Bert De Brabandere and Xu Jia and Tinne Tuytelaars and Luc Van Gool},
  journal={CoRR},
  year={2016},
  volume={abs/1605.09673}
}
In a traditional convolutional layer, the learned filters stay fixed after training. In contrast, we introduce a new framework, the Dynamic Filter Network, where filters are generated dynamically conditioned on an input. We show that this architecture is a powerful one, with increased flexibility thanks to its adaptive nature, yet without an excessive increase in the number of model parameters. A wide variety of filtering operations can be learned this way, including local spatial… CONTINUE READING
Recent Discussions
This paper has been referenced on Twitter 15 times over the past 90 days. VIEW TWEETS