Dataset Groups Activity Stream 300VW The proposed method incorporates a backbone replaceable network as the basic facial landmark detector. BibTex: @dataset{Xu_Sun_and_Zhenfeng_Fan_and_Zihao_Zhang_and_Yingjie_Guo_and_Shihong_Xia_2024, abstract = {The proposed method incorporates a backbone replaceable network as the basic facial landmark detector.}, author = {Xu Sun and Zhenfeng Fan and Zihao Zhang and Yingjie Guo and Shihong Xia}, doi = {10.57702/9x1vc3mn}, institution = {No Organization}, keyword = {'Facial Detection', 'Facial Landmarks', 'face alignment', 'stability', 'video-oriented methods'}, month = {dec}, publisher = {TIB}, title = {300VW}, url = {https://service.tib.eu/ldmservice/dataset/300vw}, year = {2024} }