HyperAI超神经

首页资讯最新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 On Potsdam 1

评估指标

Accuracy

评测结果

各个模型在此基准测试上的表现结果

模型名称	Accuracy	Paper Title	Repository
HP	82.4	Leveraging Hidden Positives for Unsupervised Semantic Segmentation
PriMaPs-EM+HP (DINO ViT-B/8)	83.3	Boosting Unsupervised Semantic Segmentation with Principal Mask Proposals
InfoSeg	-	InfoSeg: Unsupervised Semantic Image Segmentation with Mutual Information Maximization	-
STEGO	77.0	Unsupervised Semantic Segmentation by Distilling Feature Correspondences
PriMaPs-EM (DINO ViT-B/8)	80.5	Boosting Unsupervised Semantic Segmentation with Principal Mask Proposals
EAGLE (DINO, ViT-B/8)	83.3	EAGLE: Eigen Aggregation Learning for Object-Centric Unsupervised Semantic Segmentation
EQUSS	82.0	Expand-and-Quantize: Unsupervised Semantic Segmentation Using High-Dimensional Space and Product Quantization	-
IIC	45.4	Invariant Information Clustering for Unsupervised Image Classification and Segmentation

0 of 8 row(s) selected.