AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Stochastic Filter Groups for Multi-Task CNNs: Learning Specialist and Generalist Convolution Kernels

Felix J.S. Bragman, Ryutaro Tanno, Sebastien Ourselin, Daniel C. Alexander, Jorge Cardoso

ICCV 2019 · Architectures and efficiency › CNN design · approach: Large and dynamic kernel design · 78 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
learn kernel allocation (SFG) instead of fixed 1/3 mask (ablation)VGG-11 multi-task · UTKFace full dataage MAE: 7.67 -> 6.00Improved
learn kernel allocation (SFG) instead of fixed 1/3 mask (ablation)VGG-11 multi-task · UTKFace full datagender accuracy (%): 89.41 -> 92.46Improved
learn grouping probabilities instead of constant p=[1/3,1/3,1/3] (ablation)VGG-11 multi-task · UTKFace full dataage MAE: 6.34 -> 6.00Improved
learn grouping probabilities instead of constant p (ablation)VGG-11 multi-task · UTKFace small dataage MAE: 9.15 -> 8.54Improved
learn grouping probabilities instead of constant p (ablation)VGG-11 multi-task · UTKFace small datagender accuracy (%): 86.01 -> 87.01Improved
use SFG instead of hard parameter sharing (ablation)VGG-11 multi-task · UTKFace full dataage MAE: 7.92 -> 6.00Improved
use SFG instead of cross-stitchVGG-11 multi-task · UTKFace small datagender accuracy (%): 83.72 -> 87.01Improved
use SFG instead of cross-stitchHighResNet multi-task · prostate CT synthesis overallPSNR: 27.86 -> 27.74No clear effect
use SFG instead of cross-stitchHighResNet multi-task · prostate segmentation overallDICE: 0.854 -> 0.852No clear effect
learn grouping (SFG) instead of constant mask (ablation)HighResNet multi-task · prostate segmentation overallDICE: 0.774 -> 0.852Improved
use hard parameter sharing multi-task instead of single-taskHighResNet · prostate segmentation overallDICE: 0.848 -> 0.829Hurt
use SFG instead of cross-stitch on bone regionsHighResNet multi-task · prostate CT synthesis bonesPSNR: 32.27 -> 32.29No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy