update paper info

open-mmlab · Dec 13, 2023 · 46d2b85 · 46d2b85
1 parent 1b95fa0
commit 46d2b85
Show file tree

Hide file tree

Showing 7 changed files with 57 additions and 35 deletions.
diff --git a/configs/body_2d_keypoint/rtmo/body7/rtmo_body7.md b/configs/body_2d_keypoint/rtmo/body7/rtmo_body7.md
@@ -1,15 +1,16 @@
 <!-- [ALGORITHM] -->
 
 <details>
-<summary align="right"><a href="https://arxiv.org/abs/2204.06806">YOLO-Pose (CVPRW'2022)</a></summary>
+<summary align="right"><a href="https://arxiv.org/abs/2312.07526">RTMO</a></summary>
 
 ```bibtex
-@inproceedings{maji2022yolo,
-  title={Yolo-pose: Enhancing yolo for multi person pose estimation using object keypoint similarity loss},
-  author={Maji, Debapriya and Nagori, Soyeb and Mathew, Manu and Poddar, Deepak},
-  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
-  pages={2637--2646},
-  year={2022}
+@misc{lu2023rtmo,
+      title={{RTMO}: Towards High-Performance One-Stage Real-Time Multi-Person Pose Estimation},
+      author={Peng Lu and Tao Jiang and Yining Li and Xiangtai Li and Kai Chen and Wenming Yang},
+      year={2023},
+      eprint={2312.07526},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV}
 }
 ```
 

diff --git a/configs/body_2d_keypoint/rtmo/body7/rtmo_body7.yml b/configs/body_2d_keypoint/rtmo/body7/rtmo_body7.yml
@@ -1,9 +1,3 @@
-Collections:
-- Name: RTMO
-  Paper:
-    Title: 'RTMO: Towards High-Performance One-Stage Real-Time Multi-Person Pose Estimation'
-    URL: https://arxiv.org/abs/2204.06806  # TODO: update the url
-  README: https://github.com/open-mmlab/mmpose/blob/main/docs/src/papers/algorithms/yolopose.md  # TODO: update the paper
 Models:
 - Config: configs/body_2d_keypoint/rtmo/body7/rtmo-s_8xb32-600e_body7-640x640.py
   In Collection: RTMO

diff --git a/configs/body_2d_keypoint/rtmo/coco/rtmo_coco.md b/configs/body_2d_keypoint/rtmo/coco/rtmo_coco.md
@@ -1,15 +1,16 @@
 <!-- [ALGORITHM] -->
 
 <details>
-<summary align="right"><a href="https://arxiv.org/abs/2204.06806">YOLO-Pose (CVPRW'2022)</a></summary>
+<summary align="right"><a href="https://arxiv.org/abs/2312.07526">RTMO</a></summary>
 
 ```bibtex
-@inproceedings{maji2022yolo,
-  title={Yolo-pose: Enhancing yolo for multi person pose estimation using object keypoint similarity loss},
-  author={Maji, Debapriya and Nagori, Soyeb and Mathew, Manu and Poddar, Deepak},
-  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
-  pages={2637--2646},
-  year={2022}
+@misc{lu2023rtmo,
+      title={{RTMO}: Towards High-Performance One-Stage Real-Time Multi-Person Pose Estimation},
+      author={Peng Lu and Tao Jiang and Yining Li and Xiangtai Li and Kai Chen and Wenming Yang},
+      year={2023},
+      eprint={2312.07526},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV}
 }
 ```
 

diff --git a/configs/body_2d_keypoint/rtmo/coco/rtmo_coco.yml b/configs/body_2d_keypoint/rtmo/coco/rtmo_coco.yml
@@ -2,8 +2,8 @@ Collections:
 - Name: RTMO
   Paper:
     Title: 'RTMO: Towards High-Performance One-Stage Real-Time Multi-Person Pose Estimation'
-    URL: https://arxiv.org/abs/2204.06806  # TODO: update the url
-  README: https://github.com/open-mmlab/mmpose/blob/main/docs/src/papers/algorithms/yolopose.md  # TODO: update the paper
+    URL: https://arxiv.org/abs/2312.07526
+  README: https://github.com/open-mmlab/mmpose/blob/main/docs/src/papers/algorithms/rtmo.md
 Models:
 - Config: configs/body_2d_keypoint/rtmo/crowdpose/rtmo-s_8xb32-700e_crowdpose-640x640.py
   In Collection: RTMO

diff --git a/configs/body_2d_keypoint/rtmo/crowdpose/rtmo_crowdpose.md b/configs/body_2d_keypoint/rtmo/crowdpose/rtmo_crowdpose.md
@@ -1,15 +1,16 @@
 <!-- [ALGORITHM] -->
 
 <details>
-<summary align="right"><a href="https://arxiv.org/abs/2204.06806">YOLO-Pose (CVPRW'2022)</a></summary>
+<summary align="right"><a href="https://arxiv.org/abs/2312.07526">RTMO</a></summary>
 
 ```bibtex
-@inproceedings{maji2022yolo,
-  title={Yolo-pose: Enhancing yolo for multi person pose estimation using object keypoint similarity loss},
-  author={Maji, Debapriya and Nagori, Soyeb and Mathew, Manu and Poddar, Deepak},
-  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
-  pages={2637--2646},
-  year={2022}
+@misc{lu2023rtmo,
+      title={{RTMO}: Towards High-Performance One-Stage Real-Time Multi-Person Pose Estimation},
+      author={Peng Lu and Tao Jiang and Yining Li and Xiangtai Li and Kai Chen and Wenming Yang},
+      year={2023},
+      eprint={2312.07526},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV}
 }
 ```
 

diff --git a/configs/body_2d_keypoint/rtmo/crowdpose/rtmo_crowdpose.yml b/configs/body_2d_keypoint/rtmo/crowdpose/rtmo_crowdpose.yml
@@ -1,9 +1,3 @@
-Collections:
-- Name: RTMO
-  Paper:
-    Title: 'RTMO: Towards High-Performance One-Stage Real-Time Multi-Person Pose Estimation'
-    URL: https://arxiv.org/abs/2204.06806  # TODO: update the url
-  README: https://github.com/open-mmlab/mmpose/blob/main/docs/src/papers/algorithms/yolopose.md  # TODO: update the paper
 Models:
 - Config: configs/body_2d_keypoint/rtmo/coco/rtmo-s_8xb32-600e_coco-640x640.py
   In Collection: RTMO

diff --git a/docs/src/papers/algorithms/rtmo.md b/docs/src/papers/algorithms/rtmo.md
@@ -0,0 +1,31 @@
+# RTMO: Towards High-Performance One-Stage Real-Time Multi-Person Pose Estimation
+
+<!-- [ALGORITHM] -->
+
+<details>
+<summary align="right"><a href="https://arxiv.org/abs/2312.07526">RTMO</a></summary>
+
+```bibtex
+@misc{lu2023rtmo,
+      title={{RTMO}: Towards High-Performance One-Stage Real-Time Multi-Person Pose Estimation},
+      author={Peng Lu and Tao Jiang and Yining Li and Xiangtai Li and Kai Chen and Wenming Yang},
+      year={2023},
+      eprint={2312.07526},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV}
+}
+```
+
+</details>
+
+## Abstract
+
+<!-- [ABSTRACT] -->
+
+Real-time multi-person pose estimation presents significant challenges in balancing speed and precision. While two-stage top-down methods slow down as the number of people in the image increases, existing one-stage methods often fail to simultaneously deliver high accuracy and real-time performance. This paper introduces RTMO, a one-stage pose estimation framework that seamlessly integrates coordinate classification by representing keypoints using dual 1-D heatmaps within the YOLO architecture, achieving accuracy comparable to top-down methods while maintaining high speed. We propose a dynamic coordinate classifier and a tailored loss function for heatmap learning, specifically designed to address the incompatibilities between coordinate classification and dense prediction models. RTMO outperforms state-of-the-art one-stage pose estimators, achieving 1.1% higher AP on COCO while operating about 9 times faster with the same backbone. Our largest model, RTMO-l, attains 74.8% AP on COCO val2017 and 141 FPS on a single V100 GPU, demonstrating its efficiency and accuracy.
+
+<!-- [IMAGE] -->
+
+<div align=center>
+<img src="https://github.com/open-mmlab/mmpose/assets/26127467/ad94c097-7d51-4b91-b885-d8605e22a0e6" height="360px" alt><br>
+</div>