AI Paper Atlas

Research map / Detection and segmentation / Weakly supervised detection and segmentation

Learning Visual Features from Large Weakly Supervised Data

Armand Joulin, Laurens van der Maaten, Allan Jabri, Nicolas Vasilache

ECCV 2016 · Detection and segmentation › Weakly supervised detection and segmentation · approach: Weakly supervised detection · 317 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
train end-to-end on Flickr instead of logistic regression on ImageNet-pretrained featuresAlexNet word predictor · Flickr 100M word prediction (K=1,000)precision@10: 8.27 -> 17.98Improved
train end-to-end on Flickr instead of ImageNet-pretrained featuresAlexNet word predictor · Flickr 100M word prediction (K=100,000)precision@10: 1.61 -> 2.56Improved
use multiclass logistic loss instead of one-versus-all logistic loss (ablation)AlexNet, K=1,000 · Flickr 100M word predictionprecision@10: 16.43 -> 17.98Improved
increase Flickr training set beyond 1 million images (ablation)AlexNet, K=1,000 · Flickr word prediction and transferprecision@10: levels out after ~50 million imagesImproved
switch from AlexNet to GoogLeNetFlickr-trained features · ImageNet (transfer)accuracy: 34.93 -> 33.62Hurt
use Flickr features instead of ImageNet featuresAlexNet · Pascal VOC 2007 (transfer)mAP: 59.8 -> 62.4Improved
use Flickr features instead of ImageNet featuresGoogLeNet · Pascal VOC 2007 (transfer)mAP: 77.1 -> 73.2Hurt
use Flickr features instead of ImageNet featuresAlexNet · Oxford Flowers (transfer)accuracy: 80.28 -> 69.72Hurt
use Flickr features instead of ImageNet featuresAlexNet · MIT SUN (transfer)accuracy: 41.40 -> 42.67Improved
use Flickr features instead of ImageNet featuresAlexNet · MIT Indoor (transfer)accuracy: 53.82 -> 53.19No clear effect
concatenate Flickr and ImageNet featuresAlexNet · MIT Indoor (transfer)accuracy: 53.82 -> 58.76Improved
concatenate Flickr and ImageNet featuresGoogLeNet · Pascal VOC 2007 (transfer)mAP: 77.1 -> 79.0Improved

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