Bottom-up Pose Estimation of Multiple Person with Bounding Box Constraint

Li, Miaopeng; Zhou, Zimeng; Li, Jie; Liu, Xinguo

doi:10.1109/icpr.2018.8546194

Cited by 28 publications

(13 citation statements)

References 26 publications

(53 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, this approach is computationally expensive for multi-person pose estimation as each pose is estimated independent of each other. Hence, the bottom-up approach ( Li et al, 2018 ) is used in this study as an alternative to the top-down approach. It involves two steps, (i) estimation of keypoints, and (ii) establishing a relation between the keypoints to generate the pose.…”

Section: Methodsmentioning

confidence: 99%

Development of a Robust, Simple, and Affordable Human Gait Analysis System Using Bottom-Up Pose Estimation With a Smartphone Camera

et al. 2022

View full text Add to dashboard Cite

Gait analysis is used in many fields such as Medical Diagnostics, Osteopathic medicine, Comparative and Sports-related biomechanics, etc. The most commonly used system for capturing gait is the advanced video camera-based passive marker system such as VICON. However, such systems are expensive, and reflective markers on subjects can be intrusive and time-consuming. Moreover, the setup of markers for certain rehabilitation patients, such as people with stroke or spinal cord injuries, could be difficult. Recently, some markerless systems were introduced to overcome the challenges of marker-based systems. However, current markerless systems have low accuracy and pose other challenges in gait analysis with people in long clothing, hiding the gait kinematics. The present work attempts to make an affordable, easy-to-use, accurate gait analysis system while addressing all the mentioned issues. The system in this study uses images from a video taken with a smartphone camera (800 × 600 pixels at an average rate of 30 frames per second). The system uses OpenPose, a 2D real-time multi-person keypoint detection technique. The system learns to associate body parts with individuals in the image using Convolutional Neural Networks (CNNs). This bottom-up system achieves high accuracy and real-time performance, regardless of the number of people in the image. The proposed system is called the “OpenPose based Markerless Gait Analysis System” (OMGait). Ankle, knee, and hip flexion/extension angle values were measured using OMGait in 16 healthy volunteers under different lighting and clothing conditions. The measured kinematic values were compared with a standard video camera based normative dataset and data from a markerless MS Kinect system. The mean absolute error value of the joint angles from the proposed system was less than 90 for different lighting conditions and less than 110 for different clothing conditions compared to the normative dataset. The proposed system is adequate in measuring the kinematic values of the ankle, knee, and hip. It also performs better than the markerless systems like MS Kinect that fail to measure the kinematics of ankle, knee, and hip joints under dark and bright light conditions and in subjects with long robe clothing.

show abstract

Section: Methodsmentioning

confidence: 99%

Development of a Robust, Simple, and Affordable Human Gait Analysis System Using Bottom-Up Pose Estimation With a Smartphone Camera

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Existing methods before the advent of deep learning have used hand-crafted features such as Histogram of Oriented Gradients (HOG) and Local Binary Pattern (LBP) to estimate human features [1][2][3]. Recently, pose estimation methods using deep learning have shown the great performance [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20]. 2D human pose estimation has been developed by two approaches, topdown and bottom-up approaches.…”

Section: D Pose Estimationmentioning

confidence: 99%

“…Estimating human pose has received considerable attention in the field of computer vision, and has made great progress with the introduction of deep learning [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20]. Pose estimation plays an important role in various human-computer interaction tasks, such as surveillance systems and autonomous driving.…”

Section: Introductionmentioning

confidence: 99%

Multi-resolution Fusion Network for Human Pose Estimation in Low-resolution Images

2022

KSII TIIS

View full text Add to dashboard Cite

2D human pose estimation still faces difficulty in low-resolution images. Most existing topdown approaches scale up the target human bonding box images to the large size and insert the scaled image into the network. Due to up-sampling, artifacts occur in the low-resolution target images, and the degraded images adversely affect the accurate estimation of the joint positions. To address this issue, we propose a multi-resolution input feature fusion network for human pose estimation. Specifically, the bounding box image of the target human is rescaled to multiple input images of various sizes, and the features extracted from the multiple images are fused in the network. Moreover, we introduce a guiding channel which induces the multi-resolution input features to alternatively affect the network according to the resolution of the target image. We conduct experiments on MS COCO dataset which is a representative dataset for 2D human pose estimation, where our method achieves superior performance compared to the strong baseline HRNet and the previous state-of-the-art methods.

show abstract

“…HUMAN4D enables research to human pose-related computer vision tasks by providing spatio-temporally aligned RGBD data from multiple views under a HW-SYNC setting, along with accurate 3D and 2D poses. Recent research efforts are devoted on various single-and multi-person pose estimation approaches, from single RGB in the wild [18], [57]- [59], depth [60], [61], multi-view RGB [23], [62] and multiview RGBD [22], [63], among others. However, the selection criteria of the methods we benchmark are to be open-source and applicable to HUMAN4D, producing baseline results for our dataset.…”

Section: Pose Estimationmentioning

confidence: 99%

HUMAN4D: A Human-Centric Multimodal Dataset for Motions and Immersive Media

et al. 2020

View full text Add to dashboard Cite

We introduce HUMAN4D, a large and multimodal 4D dataset that contains a variety of human activities simultaneously captured by a professional marker-based MoCap, a volumetric capture and an audio recording system. By capturing 2 female and 2 male professional actors performing various fullbody movements and expressions, HUMAN4D provides a diverse set of motions and poses encountered as part of single-and multi-person daily, physical and social activities (jumping, dancing, etc.), along with multi-RGBD (mRGBD), volumetric and audio data. Despite the existence of multi-view color datasets captured with the use of hardware (HW) synchronization, to the best of our knowledge, HUMAN4D is the first and only public resource that provides volumetric depth maps with high synchronization precision due to the use of intra-and inter-sensor HW-SYNC. Moreover, a spatio-temporally aligned scanned and rigged 3D character complements HUMAN4D to enable joint research on time-varying and highquality dynamic meshes. We provide evaluation baselines by benchmarking HUMAN4D with state-of-theart human pose estimation and 3D compression methods. We apply OpenPose and AlphaPose reaching 70.02% and 82.95% mAP PCKh-0.5 on single-and 68.48% and 73.94% mAP PCKh-0.5 on two-person 2D pose estimation, respectively. In 3D pose, a recent multi-view approach named Learnable Triangulation, achieves 80.26% mAP PCK3D-10cm. For 3D compression, we benchmark Draco, Corto and CWIPC open-source 3D codecs, respecting online encoding and steady bit-rates between 7-155 and 2-90 Mbps for mesh-and pointbased volumetric video, respectively. Qualitative and quantitative visual comparison between mesh-based volumetric data reconstructed in different qualities and captured RGB, showcases the available options with respect to 4D representations. HUMAN4D is introduced to enable joint research on spatio-temporally aligned pose, volumetric, mRGBD and audio data cues. The dataset and its code are available online.

show abstract

Bottom-up Pose Estimation of Multiple Person with Bounding Box Constraint

Cited by 28 publications

References 26 publications

Development of a Robust, Simple, and Affordable Human Gait Analysis System Using Bottom-Up Pose Estimation With a Smartphone Camera

Development of a Robust, Simple, and Affordable Human Gait Analysis System Using Bottom-Up Pose Estimation With a Smartphone Camera

Multi-resolution Fusion Network for Human Pose Estimation in Low-resolution Images

HUMAN4D: A Human-Centric Multimodal Dataset for Motions and Immersive Media

Contact Info

Product

Resources

About