AI Paper Atlas

Research map / Detection and segmentation / CNN features

Multiple Saliency and Channel Sensitivity Network for Aggregated Convolutional Feature

Xuanlu Xiang, Zhipeng Wang, Zhicheng Zhao, Fei Su

AAAI 2019 · Detection and segmentation › CNN features · approach: Local features and context in CNNs · 4 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
replace SSW with Gram-CS channel weighting (off-the-shelf) (ablation)SW weighting · Oxford5KmAP: 70.8 -> 71.4Improved
replace SSW with Gram-CS channel weighting (off-the-shelf) (ablation)SW weighting · Paris6KmAP: 79.7 -> 81.5Improved
replace SSW with Gram-CS channel weighting (supervised) (ablation)SW weighting · Oxford5KmAP: 78.7 -> 80.6Improved
replace SSW with Gram-CS channel weighting (supervised) (ablation)SW weighting · Paris6KmAP: 84.9 -> 87.1Improved
multi-scale instead of single-scale input (ablation)MSB+Gram-CS · Oxford5KmAP: 81.5 -> 86.1Improved
multi-scale instead of single-scale input (ablation)MSB+Gram-CS · Paris6KmAP: 86.5 -> 90.1Improved
increase saliency masks from 1 to 5 (ablation)MSCNet supervised · Oxford5KmAP: 82.5 -> 86.1Improved
increase saliency masks from 5 to 10 (ablation)MSCNet supervised · Oxford5KmAP: 86.1 -> 83.0Hurt
increase saliency masks from 5 to 10 (ablation)MSCNet supervised · Paris6KmAP: 90.1 -> 87.4Hurt
add NGD loss to multiple saliency block (ablation)MSCNet SU-STA1 · Oxford5KmAPImproved
supervised vs unsupervised training on Holidays (ablation)MSCNet · HolidaysmAP: 85.4 -> 90.2Mixed
add query expansion (QE)MSCNet (5-512) supervised · Paris6KmAP: 90.1 -> 93.1Improved

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