Research map / 3D vision / Human pose estimation
DETRPose: Real-Time End-to-End Multi-Person Pose Estimation via Modified Transformer Decoder and Novel Denoising Keypoints
Sebastian Janampa, Marios Pattichis
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add denoising keypoints (DK) alone (ablation) | DETRPose-L, 60 queries · COCO val | AP: 68.9 -> 70.0 | Improved |
| add KSVF loss alone (ablation) | DETRPose-L, 60 queries · COCO val | AP: 68.9 -> 69.9 | Improved |
| add Pose-LQE alone (ablation) | DETRPose-L, 60 queries · COCO val | AP: 68.9 -> 69.5 | Improved |
| add all of KSVF, Pose-LQE and denoising keypoints (ablation) | DETRPose-L, 60 queries · COCO val | AP: 68.9 -> 71.5 | Improved |
| add denoising keypoints on top of Pose-LQE (ablation) | DETRPose-L, 60 queries · COCO val | AP: 69.5 -> 69.6 | Mixed |
| add denoising keypoints on top of KSVF (ablation) | DETRPose-L, 60 queries · COCO val | AP: 69.9 -> 70.9 | Improved |
| reduce instance queries from 100 to 60 (ablation) | RT-GroupPose (DETRPose-L) · COCO val | AP: 69.9 -> 68.9 | Hurt |
| initialize with D-FINE COCO-pretrained weights instead of none (ablation) | DETRPose-L · COCO val | AP: 71.5 -> 71.1 | Hurt |
| initialize with D-FINE Object365 weights instead of none (ablation) | DETRPose-L · COCO val | AP: 71.5 -> 72.7 | Improved |
| use official COCO kappa values instead of constant kappa (box area) (ablation) | denoising keypoints, DETRPose-L · COCO val | AP: 71.1 -> 71.2 | Improved |
| compute area from masks instead of box area (constant kappa) (ablation) | denoising keypoints, DETRPose-L · COCO val | AP: 71.1 -> 71.3 | Improved |
| use both COCO kappa values and mask area (ablation) | denoising keypoints, DETRPose-L · COCO val | AP: 71.1 -> 71.5 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- GroupPose — Group Pose: A Simple Baseline for End-to-End Multi-Person Pose Estimation (ICCV 2023) · better on 1 of 2 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
- Group Pose: A Simple Baseline for End-to-End Multi-Person Pose Estimation (ICCV 2023 · similarity 0.90)
- End-to-End Multi-Person Pose Estimation With Transformers (CVPR 2022 · similarity 0.89)
- DiffusionRegPose: Enhancing Multi-Person Pose Estimation using a Diffusion-Based End-to-End Regression Approach (CVPR 2024 · similarity 0.89)
- End-to-End Multi-Person Pose Estimation with Pose-Aware Video Transformer (AAAI 2026 · similarity 0.88)
- Shape-Aware Multi-Person Pose Estimation From Multi-View Images (ICCV 2021 · similarity 0.88)
- HiPoser: 3D Human Pose Estimation with Hierarchical Shared Learning at Parts-Level Using Inertial Measurement Units (AAAI 2025 · similarity 0.87)
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.
