AI Paper Atlas

Research map / Detection and segmentation / CNN features

ThunderNet: Towards Real-Time Generic Object Detection on Mobile Devices

Zheng Qin, Zeming Li, Zhaoning Zhang, Yiping Bao, Gang Yu, Yuxing Peng, Jian Sun

ICCV 2019 · Detection and segmentation › CNN features · approach: Region-based and query-based detectors · 338 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 Context Enhancement Module (ablation)compressed Light-Head R-CNN with SNet146 · COCO test-devAP: 21.6 -> 23.3Improved
add Spatial Attention Module (without CEM) (ablation)compressed Light-Head R-CNN with SNet146 · COCO test-devAP: 21.6 -> 23.0Improved
add both CEM and SAM (ablation)ThunderNet SNet146 · COCO test-devAP: 21.9 -> 23.7Improved
replace 3x3 conv in RPN with 5x5 depthwise plus 1x1 conv (ablation)Light-Head R-CNN SNet146 · COCO test-devAP: 21.9 -> 21.8No clear effect
halve fc layer in R-CNN subnet to 1024-d (ablation)Light-Head R-CNN SNet146 · COCO test-devAP: 21.8 -> 21.6No clear effect
use 5x5 instead of 3x3 depthwise convolutions (ablation)SNet146 · COCO test-devAP: 22.7 -> 23.7Improved
use 5x5 depthwise instead of double 3x3 depthwise (ablation)SNet146 · COCO test-devAP: 23.3 -> 23.7Improved
add 1024-d Conv5 to SNet146 (ablation)SNet146 · COCO test-devAP: 23.7 -> 23.2Hurt
remove Conv5 from SNet49 (ablation)SNet49 · COCO test-devAP: 19.2 -> 18.2Hurt
widen Conv5 of SNet49 to 1024-d (ablation)SNet49 · COCO test-devAP: 19.2 -> 18.8Hurt
match input resolution to backbone size (SNet146 at 320 instead of 224) (ablation)ThunderNet · COCO test-devAP: 18.7 -> 23.7Improved
use large-backbone-small-head instead of small-backbone-large-head (ablation)ThunderNet · COCO test-devAP: 20.2 -> 23.6Improved

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