個人簡介
現任上海師範大學信息與機電工程學院副院長,上海市重點基地“上海市中國小線上教育研究中心”副主任,學科帶頭人。
CCF高級會員,IEEE高級會員,CCF多媒體專委會委員,中國圖象圖形學會多媒體專委會委員,上海市圖像圖形學會理事,《中國圖象圖形學報》編委。
2011年,江西省高校中青年骨幹教師。
2013年,南昌市科技創新領軍人才“洪城特聘專家”。
2014年,江西省青年科學家培養對象。
2015~2018年,江西省重點學科“計算機科學與技術”學科帶頭人,
江西財經大學“圖像處理與智慧型信息處理”創新團隊負責人、智慧型信息處理研究所所長,江西省數字媒體重點實驗室副主任。
曾獲原機械工業部“部級優秀畢業生”、中國科學院“院長優秀獎”、江西財經大學“科研十強”、“十大優秀研究生導師”等榮譽。
主持國家科技支撐計畫子專題、國家自然科學基金項目、中國博士後基金、國家重點實驗室開放課題、江西省自然科學基金、江西省高校科技落地計畫項目、江西省教育廳科技項目等項目。在IEEE Transactions on Image Processing、IEEE Transactions on Biomedical Engineering、Pattern Recognition、Fire Safety Journal、Digital Signal Processing、Information Sciences、Neurocomputing、Computer Vision and Image Understanding、計算機學報等上發表論文90多篇。其中,SCI/EI收錄50多篇,JCR Q1區以上20多篇。授權國內發明專利5項。
積極開展產學研合作。例如,已經成功為江西科泰華軟體公司(http://www.jxkth.com)解決人臉識別、圖像分割、人像摳圖和美顏等算法難題,這些算法已經套用在自助拍照機等產品上,並研發出基於Windows、Android和iOS系統的證件照摳圖算法模組,相關產品已在在全國各地公安派出所套用,產生很好的社會效應和幾千萬元的經濟效益。採用深度學習算法,提出一種專門的端對端人像摳圖網路模型,實現了快速、準確的全自動人像摳圖智慧型算法,並在二代身份證自助拍照系統、駕駛證自助拍照系統等廣泛套用。
科研成果
在視頻煙霧事件檢測的方面,提出基於AdaBoost算法和分區不變性特徵的視頻煙霧檢測等方法,並採用Visual C++開發出一套視頻煙霧檢測系統。為了方便探索新的視頻煙霧檢測算法,我們拍攝了幾組煙霧和非煙霧視頻,並採用手工方法建立了幾套煙霧和非煙霧靜態圖像庫。
積極開展產學研合作。例如,已經成功為公司解決人臉識別、圖像分割、人像摳圖和美顏等算法難題,這些算法已經套用在自助拍照機等產品上,並研發出基於Windows、Android和iOS系統的證件照摳圖算法模組,相關產品已在在全國各地公安派出所套用,產生很好的社會效應和幾千萬元的經濟效益。採用深度學習算法,提出一種專門的端對端人像摳圖網路模型,實現了快速、準確的全自動人像摳圖智慧型算法,並在二代身份證自助拍照系統、駕駛證自助拍照系統等廣泛套用。
代表性論著
[1]. FeiniuYuan, Lin Zhang, Xue Xia,Qinghua Huang, Xuelong Li, "A Gated Recurrent Network with DualClassification Assistance for Smoke Semantic Segmentation," IEEETransactions on Image Processing, vol. 30, 2021, pp. 4409-4422. (中科院1區,影響因子9.340)
[2]. FeiniuYuan, Lin Zhang, Xue Xia, Qinghua Huang, Xuelong Li," A Wave-shaped Deep Neural Network for Smoke Density Estimation," IEEE Transactions on Image Processing,vol. 29, 2020, pp. 2301-2313. (中科院1區,影響因子9.340)
[3]. FeiniuYuan, Kai-Hsiang Chuang, and Jimin Liu. A VariationalSurface Deformation and Subdivision Based Modeling Framework for Noisy andSmall n-Furcated Tube-Like Structures. IEEETransactions on Biomedical Engineering, vol. 60, no. 6, pp. 1589-1598, Jun.2013. (中科院2區,影響因子4.424)
[4]. FeiniuYuan, Yanling Chi, Su Huang, and Jimin Liu. Modelingn-Furcated Liver vessels From a 3D Segmented Volume Using Hole Making andSubdivision Methods. IEEE Transactionson Biomedical Engineering, vol. 59, no. 2, pp. 552-561, Feb. 2012. (中科院2區,影響因子4.424)
[5]. FeiniuYuan. A double mapping framework for extraction ofshape-invariant features based on multi-scale partitions with Adaboost forvideo smoke detection. PatternRecognition, vol. 45, no. 12, pp. 4326-4336, Dec. 2012. (中科院1區,影響因子7.196)
發表的部分論文
[6]. Feiniu Yuan,Yu Zhou, Xue Xia, Xueming Qian, Jian Huang, “A Confidence Prior for ImageDehazing,” Pattern Recognition, vol. 119, 2021, pp. 1-16. (中科院1區,影響因子7.196)
[7]. FeiniuYuan, Yu Zhou, Xue Xia, Jinting Shi, Yuming Fang,Xueming Qian “Image dehazing based on a transmission fusion strategy byautomatic image matting,” ComputerVision and Image Understanding, vol. 194, 2020, pp. 1-11. (CCF 推薦的B類期刊,影響因子3.121,SCI期刊)
[8]. YumingFang, Guanqun Ding, Wenying Wen, Feiniu Yuan, Yong Yang, Zhijun Fang,Weisi Lin, "Salient Object Detection by Spatiotemporal and SemanticFeatures in Real-Time Video Processing Systems," IEEE Transactions on Industrial Electronics, vol. 67, no. 11, 2020, pp. 9893-9903.(中科院1區,影響因子7.515)
[9]. QinghuaHuang, Yonghao Huang, Yaozhong Luo, FeiniuYuan, Xuelong Li, "Segmentation of breast ultrasound image withsemantic classification of superpixels," Medical Image Analysis, vol. 61, 2020, pp. 1-13. (中科院1區,影響因子11.148)
[10]. Qinghua Huang, Fengxin Pan, Wei Li, Feiniu Yuan, Hangtong Hu, Jinhua Huang,Jie Yu, Wei Wang, “Differential Diagnosis of Atypical HepatocellularCarcinoma in Contrast-Enhanced Ultrasound Using Spatio-Temporal DiagnosticSemantics, ” IEEE Journal of Biomedicaland Health Informatics , vol. 24, no. 10, 2020, pp. 2860 - 2869. (中科院2區,影響因子5.223)
[11]. Feiniu Yuan, Gang Li, Xue Xia, Lin Zhang, Jinting Shi, “Multi-layer Gabor Filtering for VisualSmoke Recognition,” Pattern Analysis andApplications, 23, 2020, pages1117–1131. (JCR Q2區,SCI)
[12]. 袁非牛, 章琳,史勁亭,夏雪,李鋼.自編碼神經網路理論及套用綜述.計算機學報,42(1), 2019: 203-230. (EI:20191706819309)
[13]. Yuming Fang, Xiaoqiang Zhang, FeiniuYuan (通訊作者),Nevrez Imamoglu, Haiwen Liu, " Video Saliency Detection by Gestalt Theory," Pattern Recognition, vol. 96,no. 12, Dec. 2019, pp. 1-11. (中科院1區,影響因子7.196)
[14]. Feiniu Yuan,Lin Zhang, Xue Xia, Boyang Wan, Qinghua Huang, Xuelong Li, “Deep SmokeSegmentation,” Neurocomputing, vol.357, 2019, pp. 248–260. (SCI期刊,JCRQ1區,中科院2區,影響因子4.438)
[15]. Feiniu Yuan, Gang Li, Xue Xia, Lin Zhang, Jinting Shi, “FusingTexture, Edge and Line Features for Smoke Recognition,” IET Image Processing, vol. 13, no. 14, 2019, pp. 2805-2812. (JCR Q1區,CCF 推薦的C類期刊,SCI期刊論文,影響因子1.995)
[16]. Feiniu Yuan,Jinting Shi, Xue Xia, Qinghua Huang, and Xuelong Li, “Co-occurrence Matching ofLocal Binary Patterns for Improving Visual Adaption and Its application toSmoke Recognition,” IET Computer Vision,Volume 13, Issue 2, March 2019, pp. 178–187. (JCR Q2區, CCF 推薦的C類期刊,SCI期刊論文,影響因子1.516)
[17]. Feiniu Yuan,Lin Zhang, Boyang Wan, Xue Xia and Jinting Shi, "Convolutional NeuralNetworks based on Multi-Scale Additive Merging Layers for Visual SmokeRecognition, " Machine Vision andApplications, vol. 30, no. 2, 2019, pp. 345–358. (CCF 推薦的C類期刊,JCR Q2區,SCI期刊論文,影響因子1.788)
[18]. FeiniuYuan, Tiantian Tang, Xue Xia, Jinting Shi, and ShuyingLi, "Dual-Encoded Features from Both Spatial and Curvelet Domains forImage Smoke Recognition, " KSIITransactions on Internet and Information Systems, vol. 13, no. 4, pp.2078-2093, 2019. (SCI期刊,JCRQ3區)
[19]. Feiniu Yuan,Xue Xia, Jinting Shi, “Holistic learning-based high-order feature descriptorfor smoke recognition” InternationalJournal of Wavelets, Multiresolution and Information Processing, vol. 17,no. 2, 2019. (SCI期刊,JCRQ3區)
[20]. Zhifang Xia, Weiling Chen, Qiaohong Li, FeiniuYuan, Chuangeng Tian, Lu Tang, "Prediction of Mobile Image Saliencyand Quality Under Cloud Computing Environment, " Digital Signal Processing, vol. 91, 2019, pp.66-76. (JCR Q1區,影響因子2.871)
[21]. Feiniu Yuan,Jinting Shi, Xue Xia, Lin Zhang, Shuying Li, "Encoding Pairwise HammingDistances of Local Binary Patterns for Visual Smoke Recognition," Computer Vision and Image Understanding,vol. 178, pp. 43-53, Jan. 2019. (JCR Q1區,CCF推薦的B類期刊,影響因子3.121,SCI期刊)
[22]. Yu Zhou, Feiniu Yuan(通訊作者), XueXia, Ya Li, "Single image haze removal by feature mapping, " ICME Workshop, Shanghai, China, July 8to 12, 2019, pp. 144-149. (CCF B類會議,EI:20193607389700)
[23]. 夏雪,袁非牛(通訊作者),章琳,楊龍箴,史勁亭,“從傳統到深度:視覺煙霧識別、檢測到分割”,中國圖象圖形學報,火情與煙霧專欄,特邀投稿,vol. 24, no. 10, 2019: 1627-1647.
[24]. 楊龍箴,袁非牛(通訊作者),楊壽淵,雷幫軍,“基於連續圖卷積模型的視頻煙霧檢測方法”,中國圖象圖形學報,火情與煙霧專欄,特邀投稿,vol. 24, no. 10, 2019: 1658-1669.
[25]. 袁非牛,李鋼,夏雪,章琳,周宇. 採用聚合Gabor核和局部二元模式的煙霧識別方法.小型微型計算機系統, vol.40, no. 4, 2019: 827-833.
[26]. 袁非牛,夏雪,李鋼,章琳,史勁亭. 面向煙霧識別與紋理分類的Gabor網路.中國圖象圖形學報,24(2), 2019: 269-281.
[27]. Feiniu Yuan,Xue Xia, Jinting Shi, Lin Zhang, Jifeng Huang, "Learning multi-scale andmulti-order features from 3D local differences for visual smokerecognition," Information Sciences,vol. 468, pp. 193-212, Nov. 2018. (JCRQ1區區,中科院1區,影響因子5.910)
[28]. Feiniu Yuan,Xue Xia, Jinting Shi, “Mixed Co-occurrence of Local Binary Patterns and Hamming-distance-basedLocal Binary Patterns,” InformationSciences, vol. 460–461, pp. 202-222, Sep. 2018. (JCR Q1區,中科院1區,影響因子5.910)
[29]. 史勁亭, 袁非牛(通訊作者),夏雪.視頻煙霧檢測研究進展. 中國圖象圖形學報,23(3) , 2018: 303-322.
[30]. Yuming Fang, Zhijun Fang, Feiniu Yuan, Yong Yang, Shouyuan Yang,Neal N. Xiong. Optimized Multi-operator Image Retargeting Based on PerceptualSimilarity Measure. IEEE Transactions onSystems, Man, and Cybernetics: Systems, Volume 47, Issue 11,pp.2956-2966, Oct. 2017. (JCR Q1, SCI,影響因子9.309)
[31]. FeiniuYuan, Jinting Shi, Xue Xia, Yuming Fang, Zhijun Fang,Tao Mei, “High-order local ternary patterns with locality preserving projectionfor smoke detection and image classification,” Information Sciences, vol. 372, pp. 225-240, Dec. 2016. (JCR Q1區,中科院1區,影響因子5.910,SCI).
[32]. FeiniuYuan, Jinting Shi, Xue Xia, Yong Yang, Yuming Fang, RuiWang. Sub Oriented Histograms of Local Binary Patterns for Smoke Detection andTexture Classification. KSIITransactions on Internet and Information Systems, vol. 10, no. 4, pp.1807-1823, Apr. 2016. (SCI期刊,JCRQ3區)
[33]. FeiniuYuan, Zhijun Fang, Shiqian Wu, Yong Yang, Yuming Fang. AReal-Time Video Smoke Detection Using Staircase Searching Based Dual ThresholdAdaBoost and Dynamic Analysis. IET ImageProcessing, vol. 9, no. 10, pp. 849-856, Oct. 2015. (JCR Q1區,CCF 推薦的C類期刊,影響因子1.995)
[34]. FeiniuYuan. Rotation and Scale Invariant Local Binary PatternBased on High Order Directional Derivatives for Texture Classification. Digital Signal Processing, vol. 26, no.3, pp. 142-152, Mar. 2014. (JCR Q1區,影響因子2.871).
[35]. Feiniu Yuan.An interactive concave volume clipping method based on GPU ray casting withboolean operation. Computing andInformatics, vol. 31, no. 3, pp. 551-571, Sep. 2012. (SCI收錄期刊論文)
[36]. FeiniuYuan. Video-based smoke detection with histogram sequenceof LBP and LBPV pyramids. Fire SafetyJournal, vol. 46, no. 3, pp. 132-139, Mar. 2011. (JCR Q1區,火災科學國際權威期刊,影響因子2.295,該論文為2011~2016年《FireSafety Journal》近5年most cited paper)Feiniu Yuan. An integrated fire detection and suppression system based on widelyavailable video surveillance. MachineVision and Applications, vol. 21, no. 6, pp. 941-948, Jun. 2010. (CCF 推薦的C類期刊,JCR Q2區,影響因子1.788,SCI收錄期刊論文)
近期主持的項目
[1].基於Gabor卷積網路的多隱層高斯過程煙霧特徵建模研究,
國家自然科學基金(61862029),項目負責人,2019-2022。
[2].煙霧流形建模及其在視頻煙霧事件檢測中的套用,國家自然科學基金(61363038),項目負責人,2014-2017。
[3].面向煙霧檢測的局部二元模式多級提升方法研究,國家自然科學基金(61063034),項目負責人,2011-2013。
[4].高原古建築火災視頻煙霧探測技術研究,國家科技支撐計畫“以布達拉宮為代表的高原古建築火災安全技術研究”子專題(2006BAK06B07),項目負責人,2006-2008。
[5]. 江西省青年科學家培養對象,江西省科學技術廳(20142BCB23014),項目負責人,2015-2017。
[6].袁非牛,南昌市科技創新領軍人才“洪城特聘專家”,人像比對套用共享服務平台開發項目,
項目負責人,2013-2014。