TY - RPRT TI - A Training-Free Framework for Open-Vocabulary Image Segmentation and Recognition with EfficientNet and CLIP AU - Ying Dai AU - Wei Yu Chen PY - 2025 UR - https://arxiv.org/abs/2510.19333 ID - 2510.19333 ER -