AI Paper Atlas

Research map / Detection and segmentation / CNN features

Deep Feature Pyramid Reconfiguration for Object Detection

Tao Kong, Fuchun Sun, Chuanqi Tan, Huaping Liu, Wenbing Huang

ECCV 2018 · Detection and segmentation › CNN features · approach: Region-based and query-based detectors · 235 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 global attention (SE block) to local reconfiguration (ablation)SSD+Local · VOC2007 SSD300mAP (%): 79.0 -> 79.6Improved
add residual term x_l (not only R) in local reconfiguration (ablation)SSD+Local · VOC2007 SSD300mAP (%): 78.6 -> 79.0Improved
use whole feature hierarchy instead of only deeper layers (ablation)SSD+Global-Local · VOC2007 SSD300mAP (%): 79.4 -> 79.6Improved
add lateral connections to SSDSSD · VOC2007 SSD300mAP (%): 77.5 -> 78.5Improved
replace lateral connections with global-local reconfigurationSSD+lateral · VOC2007 SSD300mAP (%): 78.5 -> 79.6Improved
replace lateral connections with global-local reconfigurationSSD+lateral · VOC2007 SSD300FPS: 37 -> 39.5Improved
add global-local reconfiguration to SSDSSD · VOC2007 SSD300FPS: 44 -> 39.5Hurt
replace FPN lateral connections with global-localFaster R-CNN ResNet-101 · VOC2007 testmAP (%): 79.8 -> 80.6Improved
replace FPN lateral connections with global-localFaster R-CNN ResNet-50 · VOC2007 testmAP (%): 78.8 -> 79.4Improved
use ResNet-101 instead of VGG-16 backbone with 512 inputSSD Ours512 · VOC2007 testmAP (%): 81.1 -> 82.4Improved
enlarge input from 300 to 512SSD Ours VGG-16 · MS COCO test-devAP 0.5:0.95 (%): 28.4 -> 31.5Improved
add global-local reconfiguration to SSD513SSD513 · MS COCO small objectsAP_s (%): 10.2 -> 14.7Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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