Abstract:
His report is to dig into compact issue in vision big data search and recognition system and is to cover work and attainments grabbed by the research team of Professor Ji Rongrong in the past two years that concern compact manifestation of visual characteristics and in-depth network compaction as regards visual terminal application. In terms of compact manifestation of visual characteristics, the report is to introduce learning of sequence-sensitive hash code by introducing big-sale non-supervision sequencing information so as to retain retrieval information in the space retaining primitive high-dimensional features. In compaction of in-depth network, it is to specify in-depth cascading compact model (serial low-matrix decomposition technology) facing specific assignments (human face and visual setting interpretation) and acceleration model (structural sparse constraint pruning technologies). The report is to introduce practical application of the aforesaid studies in such visual products as Tencent, Didi, Huawei and the like.
Personal introduction:
Ji Rongrong, special term professor and doctorate supervisor in Xiamen University, Distinguished Young Scholars of the National Science Fund in 2014, top-notched youthful talent in State Ten Thousand Talents Program in 2016. His major research orientations is on computer-based vision and multimedia technologies. He has run over 100 dissertations in SCI periodicals including around 80 articles in ACM and IEEE and over 70 long articles in CCFA international sessions. His dissertations has been quoted for around 8000 times and over 2300 times respectively via Google Scholar and SCI. 12 of his dissertations knock into ESI highly-quoted/hotspot essays. He presided over joint key programs of NSFC, front-edge items for strategies in the military commission and technology commission, planned topics/sub-topics of state key research and development planning. He has been honored Microsoft Scholar Award for 2007, Award for Best Dissertation on ACM Multimedia in 2011, Outstanding Doctorate Paper in HIT in 2012, second prize of provincial natural science in 2015, first prize for technological invention by the Ministry of Education in 2016 and first prize for technological progress in Fujian in 2018.