检索结果-内蒙古大学图书馆

3D human pose and shape estimation with dense correspondence from a single depth image

VISUAL COMPUTER 2023年第1期39卷 429-441页

作者： Wang, Kangkan Zhang, Guofeng Yang, Jian Nanjing Univ Sci & Technol Sch Comp Sci & Engn Key Lab Intelligent Percept & Syst High Dimens In Minist EducPCA Lab Nanjing 210094 Peoples R China Nanjing Univ Sci & Technol Sch Comp Sci & Engn Jiangsu Key Lab Image & Video Understanding Socia Nanjing 210094 Peoples R China Zhejiang Univ State Key Lab CAD&CG Zijingang Campus Hangzhou 310058 Peoples R China

We propose a novel approach to estimate the 3D pose and shape of human bodies with dense correspondence from a single depth image. In contrast to most current 3D body model recovery methods from depth images that employ motion information of depth sequences to compute point correspondences, we reconstruct 3D human body models from a single depth image by combining the correspondence learning and the parametric model fitting. Specifically, a novel multi-view coarse-to-fine correspondence network is proposed by projecting a 3D template model into multi-view depth images. The proposed correspondence network can predict 2D flows of the input depth relative to each projected depth in a coarse-to-fine manner. The predicted multi-view flows are then aggregated to establish accurate dense point correspondences between the 3D template and the input depth with the known 3D-to-2D projection. Based on the learnt correspondences, the 3D human pose and shape represented by a parametric 3D body model are recovered through a model fitting method that incorporates an adversarial prior. We conduct extensive experiments on SURREAL, Human3.6M, DFAUST, and real depth data of human bodies. The experimental results demonstrate that the proposed method outperforms the state-of-the-art methods in terms of reconstruction accuracy.

关键词： 3D human pose and shape Dense correspondence 3D model fitting Depth image Deep learning

来源：评论

学校读者我要写书评

暂无评论

MGP-Net: Margin-Global Information Optimization-Prototype Network for Few-Shot Ancient Inscriptions Classification 12th

MGP-Net: Margin-Global Information Optimization-Prototype Ne...

引用

12th International Conference on Image and Graphics, ICIG 2023

作者： Zhu, Jiaxin He, Xuelei Liu, Haonan He, Xiaowei Xi’an Key Lab of Radiomics and Intelligent Perception School of Information Sciences and Technology Northwest University Shaanxi Xi’an China

ISBN: (纸本)9783031463105

This article focuses on the challenge of classifying bronze inscription rubbings, which have a limited number of samples and diverse characteristics. Traditional classification methods have failed to produce satisfactory results. With the emergence of meta-learning, few-shot image classification has become a popular research topic. This approach allows a classifier to recognize datasets outside the training set and complete classification with only a small number of samples. However, due to the existence of multiple categories in the ancient inscription dataset and the tendency for overfitting, existing prototype network structures have not achieved satisfactory prediction accuracy on ancient inscription datasets. To address this challenge, we propose two strategies. The first strategy is the Margin Prototype (MP), which expands the distribution of different class prototypes during the softmax operation. The second strategy is the global information optimization strategy (GioP), which reverses the prediction of support set samples to obtain more representative prototypes. Our proposed method achieves better accuracy without adding new parameters to the model. The end-to-end structural pattern remains unchanged. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

FeverNet: Enabling accurate and robust remote fever screening

引用

PATTERN RECOGNITION 2024年 156卷

作者： Yan, Mengkai Qian, Jianjun Shao, Hang Luo, Lei Yang, Jian Minist Educ PCA Lab China Jiangsu Key Lab Image & Video Understanding Key Lab Intelligent Percept & Syst High Dimens Inf Beijing Peoples R China

Remote human fever screening via thermal infrared imaging helps reduce the risk of respiratory disease transmission and plays an important role in public health monitoring. However, the accuracy of such systems often falls prey to variations in measurement distance and environment temperature. Most previous methods tend to employ sensors to overcome these variations, which are expensive schemes and have limited performance improvement. To address above problems, this paper presents a novel and robust remote fever screening framework named FeverNet. Specifically, FeverNet introduces depth estimation network and temperature distribution constraints across time periods to reduce the influence of distance variations and environment temperature changes. The fever attention module is thus proposed to enhance feature representation and expand the difference between fever faces and normal ones. In addition, we provide the Extended Thermal Infrared Face dataset (ETIF), which further gives visible images (paired with thermal infrared images) for depth estimation and improve the fever face generated method based on the maximum temperature of the face. Extensive experiments on ETIF demonstrate the advantages of our FeverNet over the state-of-the-art methods.

关键词： Fever screening Thermal infrared image Neural network Public health surveillance

来源：评论

学校读者我要写书评

暂无评论

Meta Auxiliary Learning for Facial Action Unit Detection

引用

IEEE TRANSACTIONS ON AFFECTIVE COMPUTING 2023年第3期14卷 2526-2538页

作者： Li, Yong Shan, Shiguang Nanjing Univ Sci & Technol PCA Lab Minist Educ Key Lab Intelligent Percept & Syst High Dimens inf Nanjing 210094 Peoples R China Nanjing Univ Sci & Technol Sch Comp Sci & Engn Jiangsu Key Lab Image & Video Understanding Social Nanjing 210094 Peoples R China Chinese Acad Sci Key Lab Intelligent Informat Proc Inst Comp Technol Beijing 100190 Peoples R China CAS Ctr Excellencein Brain Sci & Intelligence Tech Shanghai 200031 Peoples R China

Despite the success of deep neural networks on facial action unit (AU) detection, better performance depends on a large number of training images with accurate AU annotations. However, labeling AU is time-consuming, expensive, and error-prone. Considering AU detection and facial expression recognition (FER) are two highly correlated tasks, and facial expression (FE) is relatively easy to annotate, we consider learning AU detection and FER in a multi-task manner. However, the performance of the AU detection task cannot be always enhanced due to the negative transfer in the multi-task scenario. To alleviate this issue, we propose a Meta Auxiliary Learning method (MAL) that automatically selects highly related FE samples by learning adaptative weights for the training FE samples in a meta learning manner. The learned sample weights alleviate the negative transfer from two aspects: 1) balance the loss of each task automatically, and 2) suppress the weights of FE samples that have large uncertainties. Experimental results on several popular AU datasets demonstrate MAL consistently improves the AU detection performance compared with the state-of-the-art multi-task and auxiliary learning methods. MAL automatically estimates adaptive weights for the auxiliary FE samples according to their semantic relevance with the primary AU detection task.

关键词： Facial action unit detection auxiliary learning meta learning

来源：评论

学校读者我要写书评

暂无评论

Center-Based Decoupled Point Cloud Registration for 6D Object Pose Estimation

Center-Based Decoupled Point Cloud Registration for 6D Objec...

引用

IEEE/CVF International Conference on Computer Vision (ICCV)

作者： Jiang, Haobo Dang, Zheng Gu, Shuo Xie, Jin Salzmann, Mathieu Yang, Jian Nanjing Univ Sci & Technol PCA Lab Key Lab Intelligent Percept & Syst High Dimens In Sch Comp Sci & EngnMinist Educ Nanjing Peoples R China Nanjing Univ Sci & Technol PCA Lab Sch Comp Sci & Engn Jiangsu Key Lab Image & Video Understanding Socia Nanjing Peoples R China Ecole Polytech Fed Lausanne CVLab Lausanne Switzerland

ISBN: (纸本)9798350307184

In this paper, we propose a novel center-based decoupled point cloud registration framework for robust 6D object pose estimation in real-world scenarios. Our method decouples the translation from the entire transformation by predicting the object center and estimating the rotation in a center- aware manner. This center offset-based translation estimation is correspondence-free, freeing us from the difficulty of constructing correspondences in challenging scenarios, thus improving robustness. To obtain reliable center predictions, we use a multi-view (bird's eye view and front view) object shape description of the source-point features, with both views jointly voting for the object center. Additionally, we propose an effective shape embedding module to augment the source features, largely completing the missing shape information due to partial scanning, thus facilitating the center prediction. With the center-aligned source and model point clouds, the rotation predictor utilizes feature similarity to establish putative correspondences for SVD-based rotation estimation. In particular, we introduce a center-aware hybrid feature descriptor with a normal correction technique to extract discriminative, partaware features for high-quality correspondence construction. Our experiments show that our method outperforms the state-of-the-art methods by a large margin on realworld datasets such as TUD-L, LINEMOD, and OccludedLINEMOD. Code is available at https://***/JiangHB/CenterReg.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Fast Proximal Policy Optimization 1

引用

6th Asian Conference on Pattern Recognition (ACPR)

作者： Zhao, Weiqi Jiang, Haobo Xie, Jin Nanjing Univ Sci & Technol Minist Educ PCA Lab Key Lab Intelligent Percept & Syst High Dimens In Nanjing Peoples R China

ISBN: (数字)9783031024443

ISBN: (纸本)9783031024443;9783031024436

Proximal policy optimization (PPO) is one of the most promising deep reinforcement learning methods and has achieved remarkable success in a variety of challenging control tasks. However, its overall updating gradient of a batch of samples may mislead the optimization of some sub-samples. It potentially reduces the sample efficiency and degrades the final decision performance. Although the minimum operation of PPO can relieve it, its slow escape speed makes it difficult to escape the wrong optimization range within the limited epochs of the minibatch update. In this paper, we propose a novel fast version of PPO named fast-PPO that replaces the original minimum operation with two accelerating operations called linear-pulling and quadratic-pulling, respectively. Both of them can increase the updating weight of the gradient for the misled samples so that the gradient of the overall object follows their expected optimization direction. Extensive experiments on classic discrete control tasks and MuJoCo based continuous control tasks verify the effectiveness of our proposed fast PPO.

关键词： Proximal Policy Optimization Escape speed Accelerating operations

来源：评论

学校读者我要写书评

暂无评论

Pre-training a Density-Aware Pose Transformer for Robust LiDAR-based 3D Human Pose Estimation

arXiv

引用

arXiv 2024年

作者： An, Xiaoqi Zhao, Lin Gong, Chen Li, Jun Yang, Jian PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information Ministry of Education School of Computer Science and Engineering Nanjing University of Science and Technology China

With the rapid development of autonomous driving, LiDAR-based 3D Human Pose Estimation (3D HPE) is becoming a research focus. However, due to the noise and sparsity of LiDAR-captured point clouds, robust human pose estimation remains challenging. Most of the existing methods use temporal information, multi-modal fusion, or SMPL optimization to correct biased results. In this work, we try to obtain sufficient information for 3D HPE only by modeling the intrinsic properties of low-quality point clouds. Hence, a simple yet powerful method is proposed, which provides insights both on modeling and augmentation of point clouds. Specifically, we first propose a concise and effective density-aware pose transformer (DAPT) to get stable keypoint representations. By using a set of joint anchors and a carefully designed exchange module, valid information is extracted from point clouds with different densities. Then 1D heatmaps are utilized to represent the precise locations of the keypoints. Secondly, a comprehensive LiDAR human synthesis and augmentation method is proposed to pre-train the model, enabling it to acquire a better human body prior. We increase the diversity of point clouds by randomly sampling human positions and orientations and by simulating occlusions through the addition of laser-level masks. Extensive experiments have been conducted on multiple datasets, including IMU-annotated LidarHuman26M, SLOPER4D, and manually annotated Waymo Open Dataset v2.0 (Waymo), HumanM3. Our method demonstrates SOTA performance in all scenarios. In particular, compared with LPFormer on Waymo, we reduce the average MPJPE by 10.0mm. Compared with PRN on SLOPER4D, we notably reduce the average MPJPE by 20.7mm. Copyright © 2024, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Heterogeneous hypergraph transformer network with cross-modal future interaction for multi-agent trajectory prediction

引用

ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE 2025年 144卷

作者： Zhou, Xiangzheng Chen, Xiaobo Yang, Jian Nanjing Univ Sci & Technol Sch Comp Sci & Engn PCA Lab Nanjing 210094 Peoples R China Nanjing Univ Sci & Technol Key Lab Intelligent Percept & Syst High Dimens Inf Minist Educ Nanjing 210094 Peoples R China Shandong Technol & Business Univ Sch Comp Sci & Technol Yantai 264005 Shandong Peoples R China

Predicting the future trajectories of traffic agents is crucial for ensuring the safety and efficiency of self-driving vehicles and navigation robots in complex environments. Accurate trajectory prediction enhances decisionmaking and collision avoidance, which are fundamental to the deployment of autonomous systems in realworld scenarios such as urban traffic, delivery logistics, and pedestrian navigation. However, this task remains challenging due to agents' diverse interactions and personalized behavior based on their categories. Moreover, most trajectory prediction models neglect the interaction between generated multimodal trajectories, limiting their practical applicability. This paper proposes a heterogeneous hypergraph transformer network with cross-modal future interaction (HHT-CFI) for multi-agent trajectory prediction to address these challenges. Specifically, the trajectory coordinates are first fed into a temporal model using transformer encoders to capture temporal features. Next, a heterogeneous hypergraph transformer (HHT) is designed to model social influence between agents of different categories, capturing historical interaction features. A cross-modal future interaction (CFI) module is then developed to enable agents to perceive multiple possible future paths of their neighbors. Finally, multimodal trajectories are generated twice, where the second predictions refine the first by utilizing cross-modal future features. Extensive experiments on four datasets demonstrate the superior performance of our model over state-of-the-art methods, showcasing its potential for real-world applications in autonomous navigation and traffic management. The code of the proposed method will be publicly available at https://github. com/zhou00NJUST/HHT-CFI.

关键词： Heterogeneous hypergraph transformer Future cross-modal interaction Multi-agent trajectory prediction Multimodal trajectory prediction

来源：评论

学校读者我要写书评

暂无评论

Modeling Cardinality in Image Hashing

引用

IEEE TRANSACTIONS ON CYBERNETICS 2023年第1期53卷 114-123页

作者： Tian, Dayong Gong, Chen Gong, Maoguo Wei, Yiwen Feng, Xiaoxuan Northwestern Polytech Univ Sch Elect & Informat Xian 710071 Peoples R China Nanjing Univ Sci & Technol PCA Lab Key Lab Intelligent Percept & Syst High Dimens In Minist EducSch Comp Sci & Engn Nanjing 20094 Peoples R China Xidian Univ Key Lab Intelligent Percept & Image Understanding Minist Educ Xian 71000 Peoples R China Xidian Univ Sch Phys & Optoelect Engn Xian 71000 Peoples R China

Cardinality constraint, namely, constraining the number of nonzero outputs of models, has been widely used in structural learning. It can be used for modeling the dependencies between multidimensional labels. In hashing, the final outputs are also binary codes, which are similar to multidimensional labels. It has been validated that estimating how many 1's in a multidimensional label vector is easier than directly predicting which elements are 1 and estimating cardinality as a prior step will improve the classification performance. Hence, in this article, we incorporate cardinality constraint into the unsupervised image hashing problem. The proposed model is divided into two steps: 1) estimating the cardinalities of hashing codes and 2) then estimating which bits are 1. Unlike multidimensional labels that are known and fixed in the training phase, the hashing codes are generally learned through an iterative method and, therefore, their cardinalities are unknown and not fixed during the learning procedure. We use a neural network as a cardinality predictor and its parameters are jointly learned with the hashing code generator, which is an autoencoder in our model. The experiments demonstrate the efficiency of our proposed method.

关键词： Training Neural networks Sparse matrices Quantization (signal) Feature extraction Optimization Matrix decomposition Approximate nearest neighbors searching binary embedding cardinality image hashing

来源：评论

学校读者我要写书评

暂无评论

OTFace: Hard Samples Guided Optimal Transport Loss for Deep Face Representation

引用

IEEE TRANSACTIONS ON MULTIMEDIA 2023年 25卷 1427-1438页

作者： Qian, Jianjun Zhu, Shumin Zhao, Chaoyu Yang, Jian Wong, Wai Keung Syst High Dimens Informat Minist Educ PCA Lab Key Lab Intelligent Percept Nanjing Jiangsu Peoples R China Nanjing Univ Sci & Technol Sch Comp Sci & Engn Jiangsu Key Lab Image & Video Understanding Social Nanjing 210094 Jiangsu Peoples R China Hong Kong Polytech Univ Sch Fash & Text Hong Kong Peoples R China Lab Artificial Intelligence Design Hong Kong Peoples R China

Face representation in the wild is extremely hard due to the large scale face variations. Some deep convolutional neural networks (CNNs) have been developed to learn discriminative feature by designing properly margin-based losses, which perform well on easy samples but fail on hard samples. Although some methods mainly adjust the weights of hard samples in training stage to improve the feature discrimination, they overlook the distribution property of feature. It is worth noting that the miss-classified hard samples may be corrected from the feature distribution view. To overcome this problem, this paper proposes the hard samples guided optimal transport (OT) loss for deep face representation, OTFace in short. OTFace aims to enhance the performance of hard samples by introducing the feature distribution discrepancy while maintaining the performance on easy samples. Specifically, we embrace triplet scheme to indicate hard sample groups in one mini-batch during training. OT is then used to characterize the distribution differences of features from the high level convolutional layer. Finally, we integrate the margin-based-softmax (e.g. ArcFace or AM-Softmax) and OT together to guide deep CNN learning. Extensive experiments were conducted on several benchmark databases. The quantitative results demonstrate the advantages of the proposed OTFace over state-of-the-art methods.

关键词： Face recognition Training Task analysis Neural networks Facial features Representation learning Generators Face representation margin based softmax Optimal transport

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：