AI Paper Atlas

Research map / Detection and segmentation / CNN features

SmartAdapt: Multi-Branch Object Detection Framework for Videos on Mobiles

Ran Xu, Fangzhou Mu, Jayoung Lee, Preeti Mukherjee, Somali Chaterji, Saurabh Bagchi, Yin Li

CVPR 2022 · Detection and segmentation › CNN features · approach: Region-based and query-based detectors · 13 citations (OpenAlex)

Check 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
Use a content-aware scheduler with the MBODF (ablation)FR+MB · ILSVRC VID validationmAP: 63.6% -> 64.1%; 67.5% -> 68.3%; 69.7% -> 69.8%; 71.0% -> 71.1%Improved
Use Oracle content-aware branch selection (ablation)FR+MB, 3,942-branch MBODF · ILSVRC VID validationmAP: 63.6% -> 71.5%; 67.5% -> 75.8%; 69.7% -> 76.3%; 71.0% -> 77.6%Improved
Use light content features for accuracy prediction (ablation)368-branch FR+MB · ILSVRC VID snippetsMSE: 0.091 -> 0.083Improved
Use light content features for accuracy prediction (ablation)368-branch FR+MB · ILSVRC VID snippetsSRC: 0.377 -> 0.385Improved
Use HoC content features for accuracy prediction (ablation)368-branch FR+MB · ILSVRC VID snippetsSRC: 0.377 -> 0.387Improved
Use HOG content features for accuracy prediction (ablation)368-branch FR+MB · ILSVRC VID snippetsRecall: 0.354 -> 0.347Hurt
Use MobileNet content features for accuracy prediction (ablation)368-branch FR+MB · ILSVRC VID snippetsMSE: 0.091 -> 0.082Improved
Use MobileNet content features for accuracy prediction (ablation)368-branch FR+MB · ILSVRC VID snippetsRecall: 0.354 -> 0.368Improved
Use a trainable MobileNet feature extractor (ablation)368-branch FR+MB · ILSVRC VID snippetsRecall: 0.354 -> 0.361Improved

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