Efficient Virtual View Selection for 3D Hand Pose Estimation

Cheng, Jian; Wan, Yanguang; Zuo, Dexin; Ma, Cuixia; Gu, Jun; Tan, Ping; Wang, Hongan; Deng, Xiaoming; Zhang, Yinda

doi:10.1609/aaai.v36i1.19919

Cited by 15 publications

(5 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Early non-parametric models aim at predicting hand joints from depth maps and point clouds (Cheng et al 2022;Deng et al 2022). With the recent advances in network architectures, non-parametric models that predict 3D hand vertices become popular.…”

Section: Non-parametric Hand Modelsmentioning

confidence: 99%

Monocular 3D Hand Mesh Recovery via Dual Noise Estimation

Li,

Lin,

Huang

et al. 2024

AAAI

View full text Add to dashboard Cite

Current parametric models have made notable progress in 3D hand pose and shape estimation. However, due to the fixed hand topology and complex hand poses, current models are hard to generate meshes that are aligned with the image well. To tackle this issue, we introduce a dual noise estimation method in this paper. Given a single-view image as input, we first adopt a baseline parametric regressor to obtain the coarse hand meshes. We assume the mesh vertices and their image-plane projections are noisy, and can be associated in a unified probabilistic model. We then learn the distributions of noise to refine mesh vertices and their projections. The refined vertices are further utilized to refine camera parameters in a closed-form manner. Consequently, our method obtains well-aligned and high-quality 3D hand meshes. Extensive experiments on the large-scale Interhand2.6M dataset demonstrate that the proposed method not only improves the performance of its baseline by more than 10% but also achieves state-of-the-art performance. Project page: https://github.com/hanhuili/DNE4Hand.

show abstract

Section: Non-parametric Hand Modelsmentioning

confidence: 99%

Monocular 3D Hand Mesh Recovery via Dual Noise Estimation

Li,

Lin,

Huang

et al. 2024

AAAI

View full text Add to dashboard Cite

show abstract

“…The following works were used HandPE. For 3D hand posture estimate from a single depth, Cheng et al (2022) suggested a virtual view selection and fusion module to solve a HandPE problem through suggesting choosing many virtual perspectives at random, combining the findings, and finding that method empirically produces reliable and accurate pose estimation. They analysed the virtual views based on the confidence of virtual views using a light‐weight network via network distillation to choose the most efficient virtual views for pose fusion.…”

Section: Pose Estimation Typesmentioning

confidence: 99%

Ultimate pose estimation: A comparative study

Hassan,

Hossain,

Elmuogy

et al. 2024

Expert Systems

View full text Add to dashboard Cite

Pose estimation is a computer vision task used to detect and estimate the pose of a person or an object in images or videos. It has some challenges that can leverage advances in computer vision research and others that require efficient solutions. In this paper, we provide a preliminary review of the state‐of‐the‐art in pose estimation, including both traditional and deep learning approaches. Also, we implement and compare the performance of Hand Pose Estimation (HandPE), which uses PoseNet architecture for hand sign problems, for an ASL dataset by using different optimizers based on 10 common evaluation metrics on different datasets. Also, we discuss some related future research directions in the field of pose estimation and explore new architectures for pose estimation types. After applying the PoseNet model, the experiment results showed that the accuracy achieved was 99.9%, 89%, 97%, 79%, and 99% for the ASL alphabet, HARPET, Yoga, Animal, and Head datasets, comparing those with common optimizers and evaluation metrics on different dataset.

show abstract

“…The estimated poses were then iteratively refined to obtain the final result. In [ 49 ], an approach was proposed that uses a CNN to estimate a 3D hand pose from a single depth image. The method generates multiple viewpoints of the hand, which are then fed to another neural network to estimate hand poses for each viewpoint.…”

Section: Related Researchmentioning

confidence: 99%

Introduction to Door Opening Type Classification Based on Human Demonstration

Šimundić

Džijan

Pejić

et al. 2023

Sensors

View full text Add to dashboard Cite

Opening doors and drawers will be an important ability for future service robots used in domestic and industrial environments. However, in recent years, the methods for opening doors and drawers have become more diverse and difficult for robots to determine and manipulate. We can divide doors into three distinct handling types: regular handles, hidden handles, and push mechanisms. While extensive research has been done on the detection and handling of regular handles, the other types of handling have not been explored as much. In this paper, we set out to classify the types of cabinet door handling types. To this end, we collect and label a dataset consisting of RGB-D images of cabinets in their natural environment. As part of the dataset, we provide images of humans demonstrating the handling of these doors. We detect the poses of human hands and then train a classifier to determine the type of cabinet door handling. With this research, we hope to provide a starting point for exploring the different types of cabinet door openings in real-world environments.

show abstract

Efficient Virtual View Selection for 3D Hand Pose Estimation

Cited by 15 publications

References 15 publications

Monocular 3D Hand Mesh Recovery via Dual Noise Estimation

Monocular 3D Hand Mesh Recovery via Dual Noise Estimation

Ultimate pose estimation: A comparative study

Introduction to Door Opening Type Classification Based on Human Demonstration

Contact Info

Product

Resources

About