__其它
1)人体姿态估计
介绍：给定包含人体的图像或视频，基于深度学习算预测图像中人体2D或3D骨骼数据，并用于人体行为识别。
论文：
Xun Wang, Yan Tian*, Xuran Zhao, Tao Yang, Judith Gelernter, Jialei Wang, Guohua Cheng, and Wei Hu, “Multi-person Pose Estimation by Mask-aware Deep Reinforcement Learning,” ACM Transactions on Multimedia Computing, Communications, and Applications, 16(3), 2020, in press. (CCF-B, IF=3.27)
Yan Tian*, Wei Hu, Hangsen Hang, Jiachen Wu*, “Densely Connected Attentional Pyramid Residual Network for Human Pose Estimation”, Neurocomputing, 347, 2019:13-23. (ZKY Q2, IF=4.43)
Yan Tian, Leonid Signal, Fernando de La Torre and Yonghua Jia, “Canonical Locality Preserving Latent Variable Model for Discriminative Pose Inference”, Image and Vision Computing, 31(3), 2013: 223-230. (SCI Q3, IF=2.67)
Yan Tian, Yuan Shi and Yong Liu, “Inferring 3D Body Pose Using Variational Semi-Parametric Regression”, IEEE International Conference on Image Processing (ICIP), Brussels, Belgium, 2011
Yan Tian, Leonid Sigal, Hernan Badino, Fernando de la Torre and Yong Liu, “Latent Gaussian Mixture Regression for Human Pose Estimation”, Asian Conference on Computer Vision (ACCV), Queenstown, New Zealand, 2010.
2）人体行为识别
论文：
Yan Tian, Yifan Cao, Wei Hu, Jiachen Wu, Chao Song, Tao Yang*, “Multi-cue Combination Network for Action-based Video Classification”, IET Computer Vision, 2019. (JCR Q4, IF=1.08)
–
3）image2text
介绍：给定图像或视频，基于深度学习和语义模型得到描述图像的句子。已授权国家发明专利。
论文：
Yan Tian, Xun Wang*, Jiachen Wu, RuiliWang, Bailin Yang, “Muilti-scale Hierarchical Residual Network for Dense Captioning,” Journal of Artificial Intelligence Research, 64, 2019: 181-196. (CCF-B, IF=2.44)
专利：
田彦，王勋，黄刚。一种基于多尺度分层残差网络的图像多字幕自动生成方法。专利号201710896059.7，授权日2020-04-28。(可转让)
–
4) 自动家具布局
给定户型信息，基于人工智能算法自动得到家具的摆放位置，并可视化输出。已申请软件著作权。
软件著作权：
田彦，来锴楠，杨柏林，面向虚拟家装的家居自动布局系统。