Visual　semantic　segmentation　is　the　basic　part　of　image　understanding　and　analysis,　and　it　is　important　for　the　extraction　of　semantic　information　regions　in　images.　It　plays　an　important　role　in　the　fields　of　automatic　driving,　medical　image　diagnosis,　natural　image　analysis,　etc.　With　the　success　of　large　language　models　in　the　field　of　natural　language　processing,　the　Segment　Anything　Model　(SAM)　visual　semantic　segmentation　based　on　hints　has　led　to　a　new　paradigm　of　semantic　segmentation　in　various　fields　of　computer　vision.　In　this　review,　we　provide　a　detailed　description　of　the　applications　of　SAM　in　real-world　scenarios,　and　discuss　the　advantages　and　limitations　of　SAM　by　summarizing　and　analyzing　the　comparative　experiments　in　the　related　literature.　Finally,　the　private　customization　of　SAM　in　different　domains　is　analyzed,　and　the　future　development　of　SAM　in　the　field　of　visual　semantic　segmentation　is　prospected.　©　2023　IEEE.