Tuning-Free Universally-Supervised Semantic Segmentation

This work presents a tuning-free semantic segmentation framework based on classifying SAM masks, which is universally applicable to various types of supervision. Initially, we utilize CLIP’s zero-shot classification ability to generate pseudo-labels or perform open-vocabulary semantic segment...

Бүрэн тодорхойлолт

Номзүйн дэлгэрэнгүй

Үндсэн зохиолчид:	Xiaobo Yang, Xiaojin Gong
Формат:	Artigo
Хэл сонгох:	Inglês
Хэвлэсэн:	IEEE 2024-01-01
Цуврал:	IEEE Access
Нөхцлүүд:	Foundation model open-vocabulary semantic segmentation semantic segmentation semi-supervised semantic segmentation vision language model weakly supervised semantic segmentation.00
Онлайн хандалт:	https://ieeexplore.ieee.org/document/10779462/
Шошгууд:	Шошго нэмэх Шошго байхгүй, Энэхүү баримтыг шошголох эхний хүн болох!