From Pixels to Graphs: Open-Vocabulary Scene Graph Generation with Vision-Language Models | IEEE Conference Publication | IEEE Xplore