The conventional stage-based signal control approach with uniform phase structure is dominantly applied at signalized intersections in China so far. Such an approach is hard to efficiently handle the mixed traffic flows consisting of unbalanced volumes of motorized and non-motorized vehicles. Together with unruly road user behavior, it has resulted in many safety issues, such as traffic conflicts between the right-turning motorized vehicles and the straight-through bicycles as well as traffic conflicts at the change of phases due to clearance failure of the bicycles. A more flexible signal control approach considering safety could be able to accommodate the mixed traffic flows prevailing at signalized intersections in China. Hence, the objective of this study is to develop a group-based signal optimization model that is capable of minimizing control delay and maximizing safety at the same time. Three unique features of the proposed model and its solution algorithm are: (1) a multi-objective genetic algorithm accounting for both delay and safety during the intergreen intervals; (2) equally treating the motorists and bicyclists in the delay and safety estimation; (3) eliminating the conflicts between the motorized vehicles and bicycles. An intersection located in Shanghai was selected as the study site and a numerical calculation was solved by using a Genetic Algorithm based Matlab program. The results showed that the proposed model is able to significantly reduce the traffic conflict intensity by 35 times and thus improve safety, while it might increase control delay remarkably.