AI Paper Atlas

Research map / Detection and segmentation / CNN features

Mixture Dense Regression for Object Detection and Human Pose Estimation

Ali Varamesh, Tinne Tuytelaars

CVPR 2020 · Detection and segmentation › CNN features · approach: Efficient DETR query design · 53 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
Replace single-mode box regression with a 3-component mixture density networkCenterNet HG · COCO-val object detectionAP: 35.9 -> 39.0Improved
Replace single-mode box regression with a 3-component mixture density networkCenterNet HG · COCO-test-dev object detectionAP: 0.401 -> 0.416Improved
Replace single-mode box regression with a 3-component mixture density networkCenterNet DLA34 · COCO-test-dev object detectionAP: 0.386 -> 0.406Improved
Replace single-mode offset regression with a 2-component mixture density networkCenterNet HG · COCO-val human pose estimationAP: 46.4 -> 52.3Improved
Replace single-mode offset regression with a 3-component mixture density networkCenterNet HG · COCO-test-dev human pose estimationAP: 55.6 -> 57.9Improved
Use 5 rather than 3 mixture components (ablation)MDN object detection · COCO-val object detectionAP: 39.0 -> 38.8No clear effect
Use more than two mixture components (ablation)MDN human pose estimation · COCO human pose estimationprecisionNo clear effect
Use more than two mixture components (ablation)MDN human pose estimation · COCO human pose estimationrecall: < 1%Improved
Give each object category a separate box prediction head (ablation)Single-mode base object detector · COCO object detectionAPNo clear effect
Choose the maximum-scoring mixture component rather than mixing component outputs (ablation)MDN dense regression at inference · COCO object detection and human pose estimationAPImproved
Raise the minimum allowed variance activation value above one (ablation)MDN variance activation · COCO-valAPNo clear effect
Use mixture regression instead of single-mode regressionDense object detection · COCO-test-dev object detectionFPS: 8.44 -> 7.47Hurt

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