Caption Unification for Multi-View Lifelogging Images Based on In-Context Learning with Heterogeneous Semantic Contents | IEEE Conference Publication | IEEE Xplore