HyperAI超神经

首页算力平台文档资讯论文教程数据集百科 SOTA LLM 模型天梯 GPU 天梯顶会

中文

HyperAI超神经

Unsupervised Semantic Segmentation With 11

评估指标

mIoU

评测结果

各个模型在此基准测试上的表现结果

		Paper Title	Repository
Trident	70.8	Harnessing Vision Foundation Models for High-Performance, Training-Free Open Vocabulary Segmentation
CLS-SEG	68.7	TagCLIP: A Local-to-Global Framework to Enhance Open-Vocabulary Multi-Label Classification of CLIP Without Training
ProxyCLIP	65.0	ProxyCLIP: Proxy Attention Improves CLIP for Open-Vocabulary Segmentation
TTD (TCL)	61.1	TTD: Text-Tag Self-Distillation Enhancing Image-Text Alignment in CLIP to Alleviate Single Tag Bias
TCL	55.0	Learning to Generate Text-grounded Mask for Open-world Semantic Segmentation from Only Image-Text Pairs
TagAlign	53.9	TagAlign: Improving Vision-Language Alignment with Multi-Tag Classification
TTD (MaskCLIP)	43.1	TTD: Text-Tag Self-Distillation Enhancing Image-Text Alignment in CLIP to Alleviate Single Tag Bias
MaskCLIP	29.3	Extract Free Dense Labels from CLIP

0 of 8 row(s) selected.

Unsupervised Semantic Segmentation With 11 | SOTA | HyperAI超神经