Human Pose Model Research Articles

Exercise quantification from single camera view markerless 3D pose estimation

Sports physiotherapists and coaches are tasked with evaluating the movement quality of athletes across the spectrum of ability and experience. However, the accuracy of visual observation is low and existing technology outside of expensive lab-based solutions has limited adoption, leading to an unmet need for an efficient and accurate means to measure static and dynamic joint angles during movement, converted to movement metrics useable by practitioners. This paper proposes a set of pose landmarks for computing frequently used joint angles as metrics of interest to sports physiotherapists and coaches in assessing common strength-building human exercise movements. It then proposes a set of rules for computing these metrics for a range of common exercises (single and double drop jumps and counter-movement jumps, deadlifts and various squats) from anatomical key-points detected using video, and evaluates the accuracy of these using a published 3D human pose model trained with ground truth data derived from VICON motion capture of common rehabilitation exercises. Results show a set of mathematically defined metrics which are derived from the chosen pose landmarks, and which are sufficient to compute the metrics for each of the exercises under consideration. Comparison to ground truth data showed that root mean square angle errors were within 10° for all exercises for the following metrics: shin angle, knee varus/valgus and left/right flexion, hip flexion and pelvic tilt, trunk angle, spinal flexion lower/upper/mid and rib flare. Larger errors (though still all within 15°) were observed for shoulder flexion and ASIS asymmetry in some exercises, notably front squats and drop-jumps. In conclusion, the contribution of this paper is that a set of sufficient key-points and associated metrics for exercise assessment from 3D human pose have been uniquely defined. Further, we found generally very good accuracy of the Strided Transformer 3D pose model in predicting these metrics for the chosen set of exercises from a single mobile device camera, when trained on a suitable set of functional exercises recorded using a VICON motion capture system. Future assessment of generalization is needed.

Research on multi-feature human pose model recognition based on one-shot learning

With the development of human-computer interaction, virtual reality, and other related fields, human posture recognition has become a hot research topic. Since the human body belongs to a non-rigid model and has time-varying characteristics, the accuracy and robustness of recognition are not ideal. Based on the KinectV2 so-matosensory camera to collect skeletal information, this paper proposes a one-shot learning model matching method based on human body angle and distance characteristics. First, feature extraction is performed on the collected bone information, and the joint point vector angle and joint point displacement are calculated and a threshold is set. Secondly, the pose to be measured is matched with the template pose, and the recognition is successful if the threshold limit is met. Experimental results show that the method can detect and recognize human poses within the defined threshold in real-time, which improves the accuracy and robustness of recognition.

View-invariant Deep Architecture for Human Action Recognition using Two-stream Motion and Shape Temporal Dynamics.

Human action Recognition for unknown views, is a challenging task. We propose a deep view-invariant human action recognition framework, which is a novel integration of two important action cues: motion and shape temporal dynamics (STD). The motion stream encapsulates the motion content of action as RGB Dynamic Images (RGB-DIs), which are generated by Approximate Rank Pooling (ARP) and processed by using finetuned InceptionV3 model. The STD stream learns long-term view-invariant shape dynamics of action using a sequence of LSTM and Bi-LSTM learning models. Human Pose Model (HPM) generates view-invariant features of structural similarity index matrix (SSIM) based key depth human pose frames. The final prediction of the action is made on the basis of three types of late fusion techniques i.e. maximum (max), average (avg) and multiply (mul), applied on individual stream scores. To validate the performance of the proposed novel framework, the experiments are performed using both cross-subject and cross-view validation schemes on three publically available benchmarks- NUCLA multi-view dataset, UWA3D-II Activity dataset and NTU RGB-D Activity dataset. Our algorithm outperforms existing state-of-the-arts significantly, which is measured in terms of recognition accuracy, receiver operating characteristic (ROC) curve and area under the curve (AUC).

An Improved Structural Human Pose Model

The human pose model is the key technology of the human-computer interaction and action analysis and understanding. In this paper, an improved structural human pose model is proposed, it integrated the capability of the structural analysis in the picture structural model and the advantage of the free view of the simple staff model, with the attribute of the structure of the model, the capability of the grammar analysis is boosted up, it is the base of the human action analysis and understanding.

An approach of wavelet transform-based human pose models for indoor video applications

The development of human pose models has become more focused on monitoring and behavioural analysis of people in many video processing applications. The goal of this paper is to develop a human pose modelling using discrete wavelet transform (DWT) for the monocular videos. The human body from the video frame is extracted using wavelet transform. In this work, the low-frequency information of the wavelet features is considered for the analysis. The pose models are developed by detecting 13 features from the human body. For the motion analysis, 12 pre-defined human poses are considered. Also, the wavelet-based Delaunay triangulation modelling (DTM) is proposed for 2D models. In this paper, the hardware system is implemented for monitoring the human poses by sending the SMS message. Finally, the performance of human pose models is computed.

Skin colour segmentation based 2D and 3D human pose modelling using Discrete Wavelet Transform

Tracking of human body and their pose models from a video data is important in the indoor surveillance applications. The objective of this paper is to develop a human pose modelling in two-dimensional and three dimensional views using wavelet based skin colour segmentation for the monocular videos. The approximation information of the Discrete Wavelet Transform (DWT) is used for segmenting the human body from the video frame and to find the skin features. The human pose modelling is proposed by detecting thirteen features from the upper body as well as in the lower body. Here, twelve human poses are considered for the analysis. In this paper, the hardware system is implemented for monitoring the human poses by sending the SMS message. Also, the efficiencies of 2D modelling and 3D modelling for the pre-defined poses are evaluated.

Human Pose Model and Block Growth Combined Crowd Segmentation

Human Pose Model and Block Growth Combined Crowd Segmentation