Hand Pose Estimation from a Single RGB-D Image

We present a method for the real-time estimation of the full 3D pose of one or more human hands using a single commodity RGB camera. Recent work in the area has displayed impressive progress using RGBD input. However, since the introduction of RGBD sensors, there has been little progress for the case of monocular color input. We capitalize on the latest advancements of deep learning, combining them with the power of generative hand pose estimation techniques to achieve real-time monocular 3D hand pose estimation in unrestricted scenarios. More specifically, given an RGB image and the relevant camera calibration information, we employ a state-of-the-art detector to localize hands. Given a crop of a hand in the image, we run the pretrained network of OpenPose for hands to estimate the 2D location of hand joints. Finally, non-linear least-squares minimization fits a 3D model of the hand to the estimated 2D joint positions, recovering the 3D hand pose. Extensive experimental results provide comparison to the state of the art as well as qualitative assessment of the method in the wild.

IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000

A novel model-based approach to 3D hand tracking from monocular video is presented. The 3D hand pose, the hand texture and the illuminant are dynamically estimated through minimization of an objective function. Derived from an inverse problem formulation, the objective function enables explicit use of temporal texture continuity and shading information, while handling important self-occlusions and time-varying illumination. The minimization is done efficiently using a quasi-Newton method, for which we provide a rigorous derivation of the objective function gradient. Particular attention is given to terms related to the change of visibility near self-occlusion boundaries that are neglected in existing formulations. To this end we introduce new occlusion forces and show that using all gradient terms greatly improves the performance of the method. Qualitative and quantitative experimental results demonstrate the potential of the approach.

Pattern Recognition, 2022

Estimating the 3D pose of a hand from a 2D image is a well-studied problem and a requirement for several real-life applications such as virtual reality, augmented reality, and hand gesture recognition. Currently, reasonable estimations can be computed from single RGB images, especially when a multi-task learning approach is used to force the system to consider the shape of the hand when its pose is determined. However, depending on the method used to represent the hand, the performance can drop considerably in real-life tasks, suggesting that stable descriptions are required to achieve satisfactory results. In this paper, we present a keypoint-based end-to-end framework for 3D hand and pose estimation and successfully apply it to the task of hand gesture recognition as a study case. Specifically, after a pre-processing step in which the images are normalized, the proposed pipeline uses a multi-task semantic feature extractor generating 2D heatmaps and hand silhouettes from RGB images, a viewpoint encoder to predict the hand and camera view parameters, a stable hand estimator to produce the 3D hand pose and shape, and a loss function to guide all of the components jointly during the learning phase. Tests were performed on a 3D pose and shape estimation benchmark dataset to assess the proposed framework, which obtained state-of-the-art performance. Our system was also evaluated on two hand-gesture recognition benchmark datasets and significantly outperformed other keypoint-based approaches, indicating that it is an effective solution that is able to generate stable 3D estimates for hand pose and shape.

Advances in Intelligent Systems and Computing, 2015

Recently, model-based approaches have produced very promising results to the problems of 3D hand tracking. The current state of the art method recovers the 3D position, orientation and 20 DOF articulation of a human hand from markerless visual observations obtained by an RGB-D sensor. Hand pose estimation is formulated as an optimization problem, seeking for the hand model parameters that minimize an objective function that quantifies the discrepancy between the appearance of hand hypotheses and the actual hand observation. The design of such a function is a complicated process that requires a lot of prior experience with the problem. In this paper we automate the definition of the objective function in such optimization problems. First, a set of relevant, candidate image features is computed. Then, given synthetic data sets with ground truth information, regression analysis is used to combine these features in an objective function that seeks to maximize optimization performance. Extensive experiments study the performance of the proposed approach based on various dataset generation strategies and feature selection techniques.

2017 International Conference on Image and Vision Computing New Zealand (IVCNZ), 2017

In this work, we demonstrate a method called Deep Hand Pose Machine(DHPM) that effectively detects the anatomical joints in the human hand based on single RGB images. Current state-of-the-art methods are able to robustly infer hand poses from RGB-D images. However, the depth map from an infrared camera does not operate well under direct sunlight. Performing hand tracking outdoors using depth sensors results in unreliable depth information and inaccurate poses. For this reason we were motivated to create this method which solely utilizes ordinary RGB image without additional depth information. Our approach adapts the pose machine algorithm, which has been used in the past to detect human body joints. We perform pose machine training on synthetic data to accurately predict the position of the joints in a real hand image.

IEEE-RAS International Conference …, 2009

ERCIM News, 2013

arXiv (Cornell University), 2021

Estimating the 3D pose of a hand from a 2D image is a well-studied problem and a requirement for several real-life applications such as virtual reality, augmented reality, and hand gesture recognition. Currently, reasonable estimations can be computed from single RGB images, especially when a multi-task learning approach is used to force the system to consider the shape of the hand when its pose is determined. However, depending on the method used to represent the hand, the performance can drop considerably in real-life tasks, suggesting that stable descriptions are required to achieve satisfactory results. In this paper, we present a keypoint-based end-to-end framework for 3D hand and pose estimation and successfully apply it to the task of hand gesture recognition as a study case. Specifically, after a pre-processing step in which the images are normalized, the proposed pipeline uses a multi-task semantic feature extractor generating 2D heatmaps and hand silhouettes from RGB images, a viewpoint encoder to predict the hand and camera view parameters, a stable hand estimator to produce the 3D hand pose and shape, and a loss function to guide all of the components jointly during the learning phase. Tests were performed on a 3D pose and shape estimation benchmark dataset to assess the proposed framework, which obtained state-of-the-art performance. Our system was also evaluated on two hand-gesture recognition benchmark datasets and significantly outperformed other keypoint-based approaches, indicating that it is an effective solution that is able to generate stable 3D estimates for hand pose and shape.

In this paper we present an approach to hand pose estimation that combines both discriminative and modelbased methods to overcome the limitations of each technique in isolation. A Randomised Decision Forests (RDF) is used to provide an initial estimate of the regions of the hand. This initial segmentation provides constraints to which a 3D model is fitted using Rigid Body Dynamics. Model fitting is guided using point to surface constraints which bind a kinematic model of the hand to the depth cloud using the segmentation of the discriminative approach. This combines the advantages of both techniques, reducing the training requirements for discriminative classification and simplifying the optimization process involved in model fitting by incorporating physical constraints from the segmentation. Our experiments on two challenging sequences show that this combined method outperforms the current stateof-the-art approach.

Proceedings of the 16th International Conference on PErvasive Technologies Related to Assistive Environments

We present a method for simultaneous 3D hand shape and pose estimation on a single RGB image frame. Specifically, our method fits the MANO 3D hand model to 2D hand keypoints. Fitting is achieved based on a novel 2D objective function that exploits anatomical joint limits, combined with shape regularization on the MANO hand model, jointly optimizing the 3D shape and pose of the hand in a single frame. In a series of quantitative experiments on wellestablished datasets annotated with ground truth, we show that it is possible to obtain reconstructions that are competitive and, in some cases, superior to existing 3D hand pose estimation approaches.