AI Paper Atlas

Research map / Detection and segmentation / CNN features

Scale-Aware Trident Networks for Object Detection

Yanghao Li, Yuntao Chen, Naiyan Wang, Zhaoxiang Zhang

ICCV 2019 · Detection and segmentation › CNN features · approach: Region-based and query-based detectors · 1,254 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 multi-branch trident blocks (ablation)Faster R-CNN ResNet-101 · COCO minivalAP: 37.9 -> 39.0Improved
add weight sharing among branches (ablation)TridentNet ResNet-101 · COCO minivalAP: 39.3 -> 40.6Improved
add scale-aware training (ablation)TridentNet ResNet-101 · COCO minivalAP: 40.3 -> 40.6Improved
add scale-aware training to multi-branch without sharing (ablation)Multi-branch ResNet-101 · COCO minivalAP: 39.0 -> 39.3Improved
full TridentNet on deformable backboneFaster R-CNN ResNet-101-Deformable · COCO minivalAP: 39.9 -> 41.8Improved
increase dilation rate from 1 to 3 (ablation)Faster R-CNN ResNet-50 C4 · COCO minivalAP small: 0.174 -> 0.162Hurt
increase dilation rate from 1 to 3 (ablation)Faster R-CNN ResNet-50 C4 · COCO minivalAP large: 0.464 -> 0.492Improved
increase branches from 1 to 3 (ablation)TridentNet ResNet-50 · COCO minivalAP: 33.2 -> 36.6Improved
increase branches from 3 to 4 (ablation)TridentNet ResNet-50 · COCO minivalAP: 36.6 -> 36.5No clear effect
place trident blocks in conv4 instead of conv2 (ablation)TridentNet ResNet-50 · COCO minivalAP: 34.1 -> 36.6Improved
widen major branch scale range to all scales (ablation)TridentNet Fast · COCO minivalAP: 37.8 -> 39.3Improved
widen all branch ranges to all scales (ablation)TridentNet Fast · COCO minivalAP: 37.8 -> 40.0Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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