Text-Guided Few-Shot Semantic Segmentation with Training-Free Multimodal Feature Matching | IEEE Conference Publication | IEEE Xplore