UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Stochastic filter groups for multi-task cnns: Learning specialist and generalist convolution kernels

Bragman, F; Tanno, R; Ourselin, S; Alexander, D; Cardoso, J; (2020) Stochastic filter groups for multi-task cnns: Learning specialist and generalist convolution kernels. In: Proceedings of 2019 IEEE/CVF International Conference on Computer Vision (ICCV). (pp. pp. 1385-1394). IEEE: Seoul, Korea. Green open access

[img]
Preview
Text
Bragman_Stochastic_Filter_Groups_for_Multi-Task_CNNs_Learning_Specialist_and_Generalist_ICCV_2019_paper.pdf - Accepted version

Download (2MB) | Preview

Abstract

The performance of multi-task learning in Convolutional Neural Networks (CNNs) hinges on the design of feature sharing between tasks within the architecture. The number of possible sharing patterns are combinatorial in the depth of the network and the number of tasks, and thus hand-crafting an architecture, purely based on the human intuitions of task relationships can be time-consuming and suboptimal. In this paper, we present a probabilistic approach to learning task-specific and shared representations in CNNs for multi-task learning. Specifically, we propose "stochastic filter groups" (SFG), a mechanism to assign convolution kernels in each layer to "specialist" and "generalist" groups, which are specific to and shared across different tasks, respectively. The SFG modules determine the connectivity between layers and the structures of task-specific and shared representations in the network. We employ variational inference to learn the posterior distribution over the possible grouping of kernels and network parameters. Experiments demonstrate the proposed method generalises across multiple tasks and shows improved performance over baseline methods.

Type: Proceedings paper
Title: Stochastic filter groups for multi-task cnns: Learning specialist and generalist convolution kernels
Event: 2019 IEEE/CVF International Conference on Computer Vision (ICCV)
ISBN-13: 9781728148038
Open access status: An open access version is available from UCL Discovery
DOI: 10.1109/ICCV.2019.00147
Publisher version: https://doi.org/10.1109/ICCV.2019.00147
Language: English
Additional information: This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions.
UCL classification: UCL
UCL > Provost and Vice Provost Offices
UCL > Provost and Vice Provost Offices > UCL BEAMS
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Computer Science
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Med Phys and Biomedical Eng
URI: https://discovery.ucl.ac.uk/id/eprint/10092831
Downloads since deposit
26Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item