The scale of deer breeding has gradually increased in recent years and better information management is necessary,which requires the identification of individual deer.In this paper,a deer face dataset is produced usin...The scale of deer breeding has gradually increased in recent years and better information management is necessary,which requires the identification of individual deer.In this paper,a deer face dataset is produced using face images obtained from different angles,and an improved residual neural network(ResNet)-based recognition model is proposed to extract the features of deer faces,which have high similarity.The model is based on ResNet-50,which reduces the depth of the model,and the network depth is only 29 layers;the model connects Squeeze-and-Excitation(SE)modules at each of the four layers where the channel changes to improve the quality of features by compressing the feature information extracted through the entire layer.A maximum pooling layer is used in the ResBlock shortcut connection to reduce the information loss caused by messages passing through the ResBlock.The Rectified Linear Unit(ReLU)activation function in the network is replaced by the Exponential Linear Unit(ELU)activation function to reduce information loss during forward propagation of the network.The preprocessed 6864 sika deer face dataset was used to train the recognition model based on SEResnet,which is demonstrated to identify individuals accurately.By setting up comparative experiments under different structures,the model reduces the amount of parameters,ensures the accuracy of the model,and improves the calculation speed of the model.Using the improved method in this paper to compare with the classical model and facial recognition models of different animals,the results show that the recognition effect of this research method is the best,with an average recognition accuracy of 97.48%.The sika deer face recognition model proposed in this study is effective.The results contribute to the practical application of animal facial recognition technology in the breeding of sika deer and other animals with few distinct facial features.展开更多
This paper presents a rheology-based approach to animate realistic face model. The dynamic and biorheological characteristics of the force member （muscles） and stressed member （face） are considered. The stressed f...This paper presents a rheology-based approach to animate realistic face model. The dynamic and biorheological characteristics of the force member （muscles） and stressed member （face） are considered. The stressed face can be modeled as viscoelastic bodies with the Hooke bodies and Newton bodies connected in a composite series-parallel manner. Then, the stress-strain relationship is derived, and the constitutive equations established. Using these constitutive equations, the face model can be animated with the force generated by muscles. Experimental results show that this method can realistically simulate the mechanical properties and motion characteristics of human face, and performance of this method is satisfactory.展开更多
Active Shape Model （ASM） is a powerful statistical tool to extract the facial features of a face image under frontal view. It mainly relies on Principle Component Analysis （PCA） to statistically model the variabil...Active Shape Model （ASM） is a powerful statistical tool to extract the facial features of a face image under frontal view. It mainly relies on Principle Component Analysis （PCA） to statistically model the variability in the training set of example shapes. Independent Component Analysis （ICA） has been proven to be more efficient to extract face features than PCA. In this paper, we combine the PCA and ICA by the consecutive strategy to form a novel ASM. Firstly, an initial model, which shows the global shape variability in the training set, is generated by the PCA-based ASM. And then, the final shape model, which contains more local characters, is established by the ICA-based ASM. Experimental results verify that the accuracy of facial feature extraction is statistically significantly improved by applying the ICA modes after the PCA modes.展开更多
目的针对替牙期儿童重度腺样体肥大(AH)与颅颌面发育异常的关联,研究无创面容测量、临床症状与AH的相关性,并构建重度AH早期诊断的预测模型。方法采用横断面研究设计,纳入2023年8月—2024年12月于上海第九人民医院就诊的6~8岁替牙期儿童...目的针对替牙期儿童重度腺样体肥大(AH)与颅颌面发育异常的关联,研究无创面容测量、临床症状与AH的相关性,并构建重度AH早期诊断的预测模型。方法采用横断面研究设计,纳入2023年8月—2024年12月于上海第九人民医院就诊的6~8岁替牙期儿童201例。收集儿童颅颌面形态参数:侧面照片颜面测量角度及比值;临床症状指标包括:呼吸模式、扁桃体肥大程度、反复扁桃体发炎史、鼻炎/哮喘患病情况、OSA-18问卷得分、中耳炎史等。按照3∶1比例将研究对象分为建模组149例,验证组52例。基于鼻内镜检查将建模组分为轻/中度AH组(n=77)与重度AH组(n=72),采用LASSO回归筛选阳性变量,构建Logistic回归预测模型,并应用ROC曲线(AUC)、校准曲线及决策曲线分析(DCA)验证模型效能。结果与轻/中度AH儿童相比,重度AH儿童呈现面凸度减小[(173.24±2.71)°vs(171.01±4.08)°,P<0.001]、唇突比值增加(0.90±0.09 vs 1.02±0.11,P<0.001)、上颌凸度增加[(30.37±6.52)°vs(35.98±7.25)°,P<0.001]等特征性面容改变。呼吸模式(OR<1,P<0.05)、慢性扁桃体炎(OR=6.035,P=0.007)、慢性鼻炎(OR=5.183,P=0.013)、哮喘(OR=14.927,P=0.002)、重度鼾症(OR=5.803,P=0.011)也与重度AH之间存在关联性。基于上述因素构建的复合预测模型在建模组AUC=0.949,验证组AUC=0.961,提示该模型具有良好的区分度和临床适用性。结论本研究发现替牙期儿童的面凸度、唇突比值、上颌凸度等面容特征是重度AH的早期敏感指标。基于此构建的面容-症状预测模型提供了一种无创、高效的筛查方法,可以有效帮助重度AH的临床前诊断及广泛筛查。展开更多
目的:研究姿势性微笑时三维面部软组织变化和对称性,并分析姿势性微笑的性别差异,同时验证姿势性微笑的一致性。方法:应用光学面部三维扫描设备获取41名成年志愿者每一张休息位和两张姿势位的面部软组织图像,其中男性16人,女性25人,年龄...目的:研究姿势性微笑时三维面部软组织变化和对称性,并分析姿势性微笑的性别差异,同时验证姿势性微笑的一致性。方法:应用光学面部三维扫描设备获取41名成年志愿者每一张休息位和两张姿势位的面部软组织图像,其中男性16人,女性25人,年龄(26.76±2.70)岁。将面部图像数据导入三维分析软件进行模型定位后应用三维可变模型(3-dimensional morphable face model method,3DMM)标定软组织特征点,选取眼部、面颊部、鼻部及口周测量指标进行软组织分析,比较两种表情状态下的面中下部软组织变化情况和对称性,并分析男女差异,同时对两次姿势性微笑的测量结果进行统计学检验。结果:与休息位相比,除鼻唇角变化量(1.45°±7.65°)差异无统计学意义外,姿势性微笑时其余软组织测量值均有改变,且眼部区域也有显著变化(P<0.001)。面下部软组织主要表现为鼻基底变宽,上下唇区域后向运动,颏部前移,唇红变窄变薄,颏唇沟变浅;姿势性微笑时不对称性以口角点[2.78(1.73,3.49)mm]、眶下中点[2.36(1.22,3.27)mm]和外眼角点[2.31(1.29,2.80)mm]最为显著,另外与休息位相比,除口角点和鼻翼基部外不对称性变化差异无统计学意义(P>0.05);姿势性微笑时,男性右眼高和下唇红深度变化量大于女性(P<0.05),外眼角点和脸颊点的不对称性增加程度较女性大(P<0.05);两次姿势性微笑的一致性较好。结论:姿势性微笑时眼部、面颊部、鼻部及口周软组织存在不同程度的变化,且口角和鼻翼基部不对称性较休息位增加;另外,姿势性微笑时面部软组织的一致性较好,可为临床面部微笑美学研究提供参考。展开更多
基金This research was supported by the Science and Technology Department of Jilin Province[20210202128NC http://kjt.jl.gov.cn]The People’s Republic of China Ministry of Science and Technology[2018YFF0213606-03 http://www.most.gov.cn]+1 种基金the Jilin Province Development and Reform Commission[2019C021 http://jldrc.jl.gov.cn]the Science and Technology Bureau of Changchun City[21ZGN27 http://kjj.changchun.gov.cn].
文摘The scale of deer breeding has gradually increased in recent years and better information management is necessary,which requires the identification of individual deer.In this paper,a deer face dataset is produced using face images obtained from different angles,and an improved residual neural network(ResNet)-based recognition model is proposed to extract the features of deer faces,which have high similarity.The model is based on ResNet-50,which reduces the depth of the model,and the network depth is only 29 layers;the model connects Squeeze-and-Excitation(SE)modules at each of the four layers where the channel changes to improve the quality of features by compressing the feature information extracted through the entire layer.A maximum pooling layer is used in the ResBlock shortcut connection to reduce the information loss caused by messages passing through the ResBlock.The Rectified Linear Unit(ReLU)activation function in the network is replaced by the Exponential Linear Unit(ELU)activation function to reduce information loss during forward propagation of the network.The preprocessed 6864 sika deer face dataset was used to train the recognition model based on SEResnet,which is demonstrated to identify individuals accurately.By setting up comparative experiments under different structures,the model reduces the amount of parameters,ensures the accuracy of the model,and improves the calculation speed of the model.Using the improved method in this paper to compare with the classical model and facial recognition models of different animals,the results show that the recognition effect of this research method is the best,with an average recognition accuracy of 97.48%.The sika deer face recognition model proposed in this study is effective.The results contribute to the practical application of animal facial recognition technology in the breeding of sika deer and other animals with few distinct facial features.
基金Project supported by the National Natural Science Foundation of China (Grant No.60772124)the Shanghai Leading Academic Discipline Project (Grant No.S30108)the Outstanding Young Teachers in University Foundation of Shanghai (Grant No.B37010708003)
文摘This paper presents a rheology-based approach to animate realistic face model. The dynamic and biorheological characteristics of the force member （muscles） and stressed member （face） are considered. The stressed face can be modeled as viscoelastic bodies with the Hooke bodies and Newton bodies connected in a composite series-parallel manner. Then, the stress-strain relationship is derived, and the constitutive equations established. Using these constitutive equations, the face model can be animated with the force generated by muscles. Experimental results show that this method can realistically simulate the mechanical properties and motion characteristics of human face, and performance of this method is satisfactory.
文摘Active Shape Model （ASM） is a powerful statistical tool to extract the facial features of a face image under frontal view. It mainly relies on Principle Component Analysis （PCA） to statistically model the variability in the training set of example shapes. Independent Component Analysis （ICA） has been proven to be more efficient to extract face features than PCA. In this paper, we combine the PCA and ICA by the consecutive strategy to form a novel ASM. Firstly, an initial model, which shows the global shape variability in the training set, is generated by the PCA-based ASM. And then, the final shape model, which contains more local characters, is established by the ICA-based ASM. Experimental results verify that the accuracy of facial feature extraction is statistically significantly improved by applying the ICA modes after the PCA modes.
文摘目的针对替牙期儿童重度腺样体肥大(AH)与颅颌面发育异常的关联,研究无创面容测量、临床症状与AH的相关性,并构建重度AH早期诊断的预测模型。方法采用横断面研究设计,纳入2023年8月—2024年12月于上海第九人民医院就诊的6~8岁替牙期儿童201例。收集儿童颅颌面形态参数:侧面照片颜面测量角度及比值;临床症状指标包括:呼吸模式、扁桃体肥大程度、反复扁桃体发炎史、鼻炎/哮喘患病情况、OSA-18问卷得分、中耳炎史等。按照3∶1比例将研究对象分为建模组149例,验证组52例。基于鼻内镜检查将建模组分为轻/中度AH组(n=77)与重度AH组(n=72),采用LASSO回归筛选阳性变量,构建Logistic回归预测模型,并应用ROC曲线(AUC)、校准曲线及决策曲线分析(DCA)验证模型效能。结果与轻/中度AH儿童相比,重度AH儿童呈现面凸度减小[(173.24±2.71)°vs(171.01±4.08)°,P<0.001]、唇突比值增加(0.90±0.09 vs 1.02±0.11,P<0.001)、上颌凸度增加[(30.37±6.52)°vs(35.98±7.25)°,P<0.001]等特征性面容改变。呼吸模式(OR<1,P<0.05)、慢性扁桃体炎(OR=6.035,P=0.007)、慢性鼻炎(OR=5.183,P=0.013)、哮喘(OR=14.927,P=0.002)、重度鼾症(OR=5.803,P=0.011)也与重度AH之间存在关联性。基于上述因素构建的复合预测模型在建模组AUC=0.949,验证组AUC=0.961,提示该模型具有良好的区分度和临床适用性。结论本研究发现替牙期儿童的面凸度、唇突比值、上颌凸度等面容特征是重度AH的早期敏感指标。基于此构建的面容-症状预测模型提供了一种无创、高效的筛查方法,可以有效帮助重度AH的临床前诊断及广泛筛查。
文摘目的:研究姿势性微笑时三维面部软组织变化和对称性,并分析姿势性微笑的性别差异,同时验证姿势性微笑的一致性。方法:应用光学面部三维扫描设备获取41名成年志愿者每一张休息位和两张姿势位的面部软组织图像,其中男性16人,女性25人,年龄(26.76±2.70)岁。将面部图像数据导入三维分析软件进行模型定位后应用三维可变模型(3-dimensional morphable face model method,3DMM)标定软组织特征点,选取眼部、面颊部、鼻部及口周测量指标进行软组织分析,比较两种表情状态下的面中下部软组织变化情况和对称性,并分析男女差异,同时对两次姿势性微笑的测量结果进行统计学检验。结果:与休息位相比,除鼻唇角变化量(1.45°±7.65°)差异无统计学意义外,姿势性微笑时其余软组织测量值均有改变,且眼部区域也有显著变化(P<0.001)。面下部软组织主要表现为鼻基底变宽,上下唇区域后向运动,颏部前移,唇红变窄变薄,颏唇沟变浅;姿势性微笑时不对称性以口角点[2.78(1.73,3.49)mm]、眶下中点[2.36(1.22,3.27)mm]和外眼角点[2.31(1.29,2.80)mm]最为显著,另外与休息位相比,除口角点和鼻翼基部外不对称性变化差异无统计学意义(P>0.05);姿势性微笑时,男性右眼高和下唇红深度变化量大于女性(P<0.05),外眼角点和脸颊点的不对称性增加程度较女性大(P<0.05);两次姿势性微笑的一致性较好。结论:姿势性微笑时眼部、面颊部、鼻部及口周软组织存在不同程度的变化,且口角和鼻翼基部不对称性较休息位增加;另外,姿势性微笑时面部软组织的一致性较好,可为临床面部微笑美学研究提供参考。
