Multi-Task Learning By A Top-Down Control Network

As the range of tasks performed by a general vision system expands, executing multiple tasks accurately and efficiently in a single network has become an important and still open problem. Recent computer vision approaches address this problem by branching networks, or by a channel-wise modulation of...

Full description

Saved in:
Bibliographic Details
Main Authors: Levi, Hila, Ullman, Shimon
Format: Conference Proceeding
Language:English
Subjects:
Computer architecture
Computer vision
Conferences
Deep learning
Image recognition
Machine vision
Modulation
Multi-task learning
Task analysis
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:As the range of tasks performed by a general vision system expands, executing multiple tasks accurately and efficiently in a single network has become an important and still open problem. Recent computer vision approaches address this problem by branching networks, or by a channel-wise modulation of the network feature-maps with task specific vectors. We present a novel architecture that uses a dedicated top-down control network to modify the activation of all the units in the main recognition network in a manner that depends on the selected task, image content, and spatial location. We show the effectiveness of our scheme by achieving significantly better results than alternative state-of-the-art approaches on four datasets. We further demonstrate our advantages in terms of task selectivity, scaling the number of tasks and interpretability.
ISSN:2381-8549
DOI:10.1109/ICIP42928.2021.9506501