HyperAI超神经

首页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

Few Shot Semantic Segmentation On Coco 20I

评估指标

Mean IoU

评测结果

各个模型在此基准测试上的表现结果

模型名称	Mean IoU	Paper Title	Repository
FPTrans (DeiT-B/16)	69.7	Feature-Proxy Transformer for Few-Shot Segmentation
HSNet (HM, ResNet-50)	65.2	HM: Hybrid Masking for Few-Shot Segmentation
DGPNet (ResNet-101)	70.1	Dense Gaussian Processes for Few-Shot Segmentation
HSNet (HM, ResNet-101)	66.5	HM: Hybrid Masking for Few-Shot Segmentation
MSDNet (ResNet-101)	73.9	MSDNet: Multi-Scale Decoder for Few-Shot Semantic Segmentation via Transformer-Guided Prototyping
VAT (HM, ResNet-50)	65.1	HM: Hybrid Masking for Few-Shot Segmentation
RPMM	49.6	Prototype Mixture Models for Few-shot Semantic Segmentation
CWT (ResNet-50)	59.5	Simpler is Better: Few-shot Semantic Segmentation with Classifier Weight Transformer
RePRI (ResNet-50)	63.1	Few-Shot Segmentation Without Meta-Learning: A Good Transductive Inference Is All You Need?
VAT + MSI (ResNet101)	69.2	MSI: Maximize Support-Set Information for Few-Shot Segmentation
DGPNet (ResNet-50)	68.9	Dense Gaussian Processes for Few-Shot Segmentation
MSDNet (ResNet-50)	72.1	MSDNet: Multi-Scale Decoder for Few-Shot Semantic Segmentation via Transformer-Guided Prototyping
FPTrans (ViT-B/16)	67.6	Feature-Proxy Transformer for Few-Shot Segmentation

0 of 13 row(s) selected.