AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Locally-Transferred Fisher Vectors for Texture Classification

Yang Song, Fan Zhang, Qing Li, Heng Huang, Lauren J. O'Donnell, Weidong Cai

ICCV 2017 · Architectures and efficiency › CNN design · approach: Filter design and trained-filter analysis · 64 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
transform FV-CNN with LFV networkFV-CNN · KTH-TIPS2accuracy (%): 81.4 -> 82.6Improved
transform FV-CNN with LFV networkFV-CNN · FMDaccuracy (%): 79.7 -> 82.1Improved
transform FV-CNN with LFV networkFV-CNN · DTDaccuracy (%): 72.4 -> 73.8Improved
fine-tune VGG-VD on texture dataset before FV encodingFV-CNN · DTDaccuracy (%): 72.4 -> 71.8Hurt
learn FV descriptor end-to-end with an FV layerFV-CNN · DTDaccuracy (%): 72.4 -> 68.2Hurt
concatenate transformed FC-CNN descriptor with LFVLFV · FMDaccuracy (%): 82.1 -> 83.5Improved
concatenate transformed FC-CNN descriptor with LFVLFV · DTDaccuracy (%): 73.8 -> 75.2Improved
concatenate transformed FC-CNN descriptor with LFVLFV · KTH-TIPS2accuracy (%): 82.6 -> 83.1Improved
use LFV instead of compact bilinear pooling at 8192 dimsCBP-TS · DTD (1/3 train, 2/3 test)accuracy (%): 67.8 -> 68.6Improved
use locally connected layer instead of fully connected layer (ablation)LFV · DTDaccuracy (%)Improved
use SVM hinge loss instead of softmax loss (ablation)LFV · KTH-TIPS2 / FMD / DTDaccuracy (%): 0.5% lower accuracy on average with softmaxImproved
share weights across every 8 filters instead of 4 (P=8) (ablation)LFV · KTH-TIPS2 / FMD / DTDaccuracy (%)Hurt

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