Facial emotion recognition is an essential and important aspect of the field of human-machine interaction.Past research on facial emotion recognition focuses on the laboratory environment.However,it faces many challen...Facial emotion recognition is an essential and important aspect of the field of human-machine interaction.Past research on facial emotion recognition focuses on the laboratory environment.However,it faces many challenges in real-world conditions,i.e.,illumination changes,large pose variations and partial or full occlusions.Those challenges lead to different face areas with different degrees of sharpness and completeness.Inspired by this fact,we focus on the authenticity of predictions generated by different<emotion,region>pairs.For example,if only the mouth areas are available and the emotion classifier predicts happiness,then there is a question of how to judge the authenticity of predictions.This problem can be converted into the contribution of different face areas to different emotions.In this paper,we divide the whole face into six areas:nose areas,mouth areas,eyes areas,nose to mouth areas,nose to eyes areas and mouth to eyes areas.To obtain more convincing results,our experiments are conducted on three different databases:facial expression recognition+(FER+),real-world affective faces database(RAF-DB)and expression in-the-wild(ExpW)dataset.Through analysis of the classification accuracy,the confusion matrix and the class activation map(CAM),we can establish convincing results.To sum up,the contributions of this paper lie in two areas:1)We visualize concerned areas of human faces in emotion recognition;2)We analyze the contribution of different face areas to different emotions in real-world conditions through experimental analysis.Our findings can be combined with findings in psychology to promote the understanding of emotional expressions.展开更多
Hester-Dendy (HD) multi-plate samplers have been widely used by state and federal government agencies for bioassessment of water quality through use of macroinvertebrate community data. To help guide remediation and r...Hester-Dendy (HD) multi-plate samplers have been widely used by state and federal government agencies for bioassessment of water quality through use of macroinvertebrate community data. To help guide remediation and restoration efforts at the Niagara River Great Lakes Area of Concern site, a multi-agency study was conducted in 2014 to assess the contribution of seven major urban tributaries on the US side of the river toward the impairment of the Niagara River. As part of this study, macroinvertebrate communities were sampled using two co-located versions of HD samplers: one version used by the New York State Department of Environmental Conservation (NYSDEC) and another by the US Environmental Protection Agency Office of Research and Development. Samplers were deployed in tributaries in highly developed watersheds with high percent impervious surface. The two sampling methods varied in terms of number and size of plates, between-plate spacing, and deployment method. Comparison of the similarity/grouping of communities with multivariate ordination techniques, Nonmetric Multidimensional Scaling and Multi-Response Permutation Procedure, showed that both methods were able to detect differences in communities at stations, despite some grouping by month and method. The indices and metrics derived from the two HD methods were found to give comparable but not identical assessments of water quality. Despite their differences, the methods were robust with respect to water quality categories derived from indices used nationally (HBI) and by NY state (BAP). For the common richness metrics, total taxa and EPT richness, there was no statistical difference between means from 3 samplings. Some metrics, especially percent tolerant collector-gatherer individuals, did show significant differences at certain stations. Indicator Species Analysis showed some taxa associated with each method. The observed community differences were thought mostly due to the difference in sampler deployment position. .展开更多
基金supported by the National Key Research & Development Plan of China (No. 2017YFB1002804)National Natural Science Foundation of China (Nos. 61425017, 61773379, 61332017, 61603390 and 61771472)the Major Program for the 325 National Social Science Fund of China (No. 13&ZD189)
文摘Facial emotion recognition is an essential and important aspect of the field of human-machine interaction.Past research on facial emotion recognition focuses on the laboratory environment.However,it faces many challenges in real-world conditions,i.e.,illumination changes,large pose variations and partial or full occlusions.Those challenges lead to different face areas with different degrees of sharpness and completeness.Inspired by this fact,we focus on the authenticity of predictions generated by different<emotion,region>pairs.For example,if only the mouth areas are available and the emotion classifier predicts happiness,then there is a question of how to judge the authenticity of predictions.This problem can be converted into the contribution of different face areas to different emotions.In this paper,we divide the whole face into six areas:nose areas,mouth areas,eyes areas,nose to mouth areas,nose to eyes areas and mouth to eyes areas.To obtain more convincing results,our experiments are conducted on three different databases:facial expression recognition+(FER+),real-world affective faces database(RAF-DB)and expression in-the-wild(ExpW)dataset.Through analysis of the classification accuracy,the confusion matrix and the class activation map(CAM),we can establish convincing results.To sum up,the contributions of this paper lie in two areas:1)We visualize concerned areas of human faces in emotion recognition;2)We analyze the contribution of different face areas to different emotions in real-world conditions through experimental analysis.Our findings can be combined with findings in psychology to promote the understanding of emotional expressions.
文摘Hester-Dendy (HD) multi-plate samplers have been widely used by state and federal government agencies for bioassessment of water quality through use of macroinvertebrate community data. To help guide remediation and restoration efforts at the Niagara River Great Lakes Area of Concern site, a multi-agency study was conducted in 2014 to assess the contribution of seven major urban tributaries on the US side of the river toward the impairment of the Niagara River. As part of this study, macroinvertebrate communities were sampled using two co-located versions of HD samplers: one version used by the New York State Department of Environmental Conservation (NYSDEC) and another by the US Environmental Protection Agency Office of Research and Development. Samplers were deployed in tributaries in highly developed watersheds with high percent impervious surface. The two sampling methods varied in terms of number and size of plates, between-plate spacing, and deployment method. Comparison of the similarity/grouping of communities with multivariate ordination techniques, Nonmetric Multidimensional Scaling and Multi-Response Permutation Procedure, showed that both methods were able to detect differences in communities at stations, despite some grouping by month and method. The indices and metrics derived from the two HD methods were found to give comparable but not identical assessments of water quality. Despite their differences, the methods were robust with respect to water quality categories derived from indices used nationally (HBI) and by NY state (BAP). For the common richness metrics, total taxa and EPT richness, there was no statistical difference between means from 3 samplings. Some metrics, especially percent tolerant collector-gatherer individuals, did show significant differences at certain stations. Indicator Species Analysis showed some taxa associated with each method. The observed community differences were thought mostly due to the difference in sampler deployment position. .
