HIGHLIGHTS
SUMMARY
The problem of hand detection is not a new study; however, the problem persists when it comes to detection in the EV datasets. Often, studies using deep learning models for 3D hand pose estimation and hand activity recognition apply third-person viewpoint datasets such as the NYU, ICVL, and MSRA datasets. During the research, the authors performed a study using Google Mediapipe (GM) for hand detection and classification on the HOI4D dataset. The results show that the pre-trained models of models have low results in hand detection and classification. The authors manually mark . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.