Loading...
Loading...

Go to the content (press return)

Semantic embeddings in deep convolutional neural networks

Author
Vilalta, A.
Type of activity
Presentation of work at congresses
Name of edition
24th European Conference on Artificial Intelligence
Date of publication
2020
Presentation's date
2020-08
Book of congress proceedings
Proceedings of the 1st Doctoral Consortium at the European Conference on Artificial Intelligence (DC-ECAI 2020)
First page
77
Last page
78
Abstract
One of the big questions in Artificial Intelligence (AI) is how can we represent knowledge. Here we focus on the field of transfer learning, specifically: how can we represent the knowledge learned by a Deep Neural Network for one specific problem, to be of use in different problems? We already found two important points: the contextualization of the knowledge learned and how a reduction of expressiveness can increase generalisation.

Participants