02/02/2021

Towards Reusable Network Components by Learning Compatible Representations

Michael Gygli, Jasper Uijlings, Vittorio Ferrari

Keywords:

Abstract Paper Similar Papers

Abstract: This paper proposes to make a first step towards compatible and hence reusable network components. Rather than training networks for different tasks independently, we adapt the training process to produce network components that are compatible across tasks. In particular, we split a network into two components, a features extractor and a target task head, and propose various approaches to accomplish compatibility between them. We systematically analyse these approaches on the task of image classification on standard datasets. We demonstrate that we can produce components which are directly compatible without any fine-tuning or compromising accuracy on the original tasks. Afterwards, we demonstrate the use of compatible components on three applications: Unsupervised domain adaptation, transferring classifiers across feature extractors with different architectures, and increasing the computational efficiency of transfer learning.

The video of this talk cannot be embedded. You can watch it here:
https://slideslive.com/38949240
(Link will open in new window)
 0
 0
 0
 0
  Share    
This is an embedded video. Talk and the respective paper are published at AAAI 2021 virtual conference. If you are one of the authors of the paper and want to manage your upload, see the question "My papertalk has been externally embedded..." in the FAQ section.

Comments

Post Comment
no comments yet
code of conduct: tbd Characters remaining: 140

Similar Papers

 0
 0
 0
 0
 4:57 
 0
 0
 0
 0
 3:18 
 0
 0
 0
 0
 5:12 
 0
 0
 0
 0
 12:02 
 0
 0
 0
 0
 5:07 
 0
 0
 0
 0
 4:52 
 0
 0
 0
 0
 9:06 
 0
 0
 0
 0
 5:03 
 0
 0
 0
 0
 3:25 
 0
 0
 0
 0
 14:13 
 0
 0
 0
 0
 4:43 
 0
 0
 0
 0
 16:15 
 0
 0
 0
 0
 3:28 
 0
 0
 0
 0
 1:00 
 0
 0
 0
 0
 14:06 
 0
 0
 0
 0
 5:26 
 0
 0
 0
 0
 0:59 
 0
 0
 0
 0
 15:58 
 0
 0
 0
 0
 3:09 
 0
 0
 0
 0
 2:32 
 0
 0
 0
 0
 5:01 
 0
 0
 0
 0
 14:47 
 0
 0
 0
 0
 7:36 
 0
 0
 0
 0
 13:56 
 0
 0
 0
 0
 4:01 
 0
 0
 0
 0
 5:04 
 0
 0
 0
 0
 0:59 
 0
 0
 0
 0
 11:46 
 2
 1
 0
 0
 13:27 
 0
 0
 0
 0
 15:41 
 0
 0
 0
 0
 13:59 
 0
 0
 0
 0
 14:36 
 0
 0
 0
 0
 1:00 
 0
 0
 0
 0
 14:43 
 0
 0
 0
 0
 12:54 
 0
 0
 0
 0
 14:19 
 0
 0
 0
 0
 16:06 
 0
 0
 0
 0
 9:52 
 0
 0
 0
 0
 14:19 
 0
 0
 0
 1
 9:03