Fig 10 - uploaded by Jingpei Lu
Content may be subject to copyright.
Qualitative results of the tool tracking for three different setups. From top to bottom raw the figures show the results of Hand-labelled SuPer Deep Keypoints, the Syntheticallylabelled SuPer Deep Keypoints, and the Optimized Keypoints. The green area shows the intersection of the rendered mask and ground-truth mask (G ∩ P ), and the red area shows the difference between the rendered mask and ground-truth mask (G ∪ P − G ∩ P ).
Source publication
Keypoint detection is an essential building block for many robotic applications like motion capture and pose estimation. Historically, keypoints are detected using uniquely engineered markers such as checkerboards, fiducials, or markers. More recently, deep learning methods have been explored as they have the ability to detect user-defined keypoint...
Contexts in source publication
Context 1
... tool tracking performance is computed by rendering a re-projected tool mask on the image frame from the estimation based on the keypoints. The IoU is computed with the groundtruth mask for evaluation. The quantitative results are shown in Fig. 9, and the qualitative results are shown in Fig. 10. The setup with Hand-labelled SuPer Deep Keypoints fails to track the tool when the tool was turning, as those hand-picked keypoints on the tool surface are occluded and humans cannot provide the label for those non-visible keypoints. However, using the optimized keypoints, the DNN makes pretty accurate predictions for non-visible ...
Context 2
... tool tracking performance is computed by rendering a re-projected tool mask on the image frame from the estimation based on the keypoints. The IoU is computed with the groundtruth mask for evaluation. The quantitative results are shown in Fig. 9, and the qualitative results are shown in Fig. 10. The setup with Hand-labelled SuPer Deep Keypoints fails to track the tool when the tool was turning, as those hand-picked keypoints on the tool surface are occluded and humans cannot provide the label for those non-visible keypoints. However, using the optimized keypoints, the DNN makes pretty accurate predictions for non-visible ...
Similar publications
The surf pop-up is a unique and challenging skill, critical to successful surfing. Hypothesizing that anthropometric characteristics of surfers influence the pop-up performance, we aimed to measure kinematics and ground-reaction forces (GRF) during a simulated pop-up motion, and to relate these variables with anthropometric characteristics. Twenty-...
