AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Increasing CNN Robustness to Occlusions by Reducing Filter Support

Elad Osherov, Michael Lindenbaum

ICCV 2017 · Detection and segmentation › Open-set and few-shot detection · approach: Object-level representation learning · 44 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
add second-central-moment regularizer on conv filtersLeNet trained on unoccluded images · CIFAR-10 (LeNet, occluded val, delta=0.1)error rate (%): 27.31 -> 23.47Improved
add group-sparsity regularizer on conv filtersLeNet trained on unoccluded images · CIFAR-10 (LeNet, occluded val, delta=0.1)error rate (%): 27.31 -> 24.20Improved
add mutually re-weighted L1 regularizer on conv filtersLeNet trained on unoccluded images · CIFAR-100 (LeNet, occluded val, delta=0.1)error rate (%): 60.17 -> 58.68Improved
add plain L1 regularization on conv+fcLeNet trained on unoccluded images · CIFAR-10 (LeNet, occluded val, delta=0.1)error rate (%): 27.31 -> 27.52No clear effect
add iterative reweighted L1 on conv+fcLeNet trained on unoccluded images · CIFAR-10 (LeNet, occluded val, delta=0.1)error rate (%): 27.31 -> 26.45Improved
also regularize first fully connected layer with second-central-moment term (ablation)RCM conv · CIFAR-10 (LeNet, occluded val, delta=0.1)error rate (%): 23.47 -> 23.00Improved
also regularize first fully connected layer with mutually re-weighted L1 (ablation)RMRWL1 conv · CIFAR-100 (LeNet, occluded val, delta=0.1)error rate (%): 58.68 -> 59.78Hurt
train on occluded examples instead of unoccludedLeNet · CIFAR-10 (LeNet, occluded val, delta=0.1)error rate (%): 27.31 -> 22.89Improved
RCM regularization vs training on occluded examplesLeNet · CIFAR-100 (LeNet, occluded val, delta=0.1)error rate (%): 56.43 -> 56.22Improved
use strong lambda tuned for occlusion, tested on unoccluded imagesRCM · CIFAR (LeNet, unoccluded val)error rate (%): increases by 0.4%Hurt
RCM with lambda tuned for unoccluded images instead of L2RCM · CIFAR (LeNet, unoccluded val)error rate (%): goes down by 1.7%Improved
add support regularization to VGG16 instead of occluded trainingVGG16 · ILSVRC12 (occluded val, delta=0.1)top-5 error (%): 16.9 -> 13.3Improved

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