HyperAI초신경
홈뉴스최신 연구 논문튜토리얼데이터셋백과사전SOTALLM 모델GPU 랭킹컨퍼런스
전체 검색
소개
HyperAI초신경

Few Shot Semantic Segmentation On Pascal 5I 1

평가 지표

Mean IoU

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름
Mean IoU
Paper TitleRepository
RePRI (CECE-T,ResNet-50)60.5Clustered-patch Element Connection for Few-shot Learning-
DGPNet (ResNet-101)64.8Dense Gaussian Processes for Few-Shot Segmentation
DCAMA (Swin-B)69.3Dense Cross-Query-and-Support Attention Weighted Mask Aggregation for Few-Shot Segmentation
GF-SAM72.1Bridge the Points: Graph-based Few-shot Segment Anything Semantically
IPRNet (ResNet-50)65.7Interclass Prototype Relation for Few-Shot Segmentation-
PGNet (ResNet-50)56.0Pyramid Graph Networks With Connection Attentions for Region-Based One-Shot Semantic Segmentation-
IMR-HSNet (ResNet-50)61.1Iterative Few-shot Semantic Segmentation from Image Label Text
ASGNet (ResNet-101)59.31Adaptive Prototype Learning and Allocation for Few-Shot Segmentation
VAT (HM, ResNet-50)65.8HM: Hybrid Masking for Few-Shot Segmentation-
APANet (ResNet-101)64APANet: Adaptive Prototypes Alignment Network for Few-Shot Semantic Segmentation-
DACM (ResNet-101)67.5Doubly Deformable Aggregation of Covariance Matrices for Few-shot Segmentation
HDMNet (VGG-16)65.1Hierarchical Dense Correlation Distillation for Few-Shot Segmentation
DACM (ResNet-50)65.7Doubly Deformable Aggregation of Covariance Matrices for Few-shot Segmentation
FECANet (VGG-16)64.3FECANet: Boosting Few-Shot Semantic Segmentation with Feature-Enhanced Context-Aware Network
HMNet (VGG-16)67.3Hybrid Mamba for Few-Shot Segmentation
PGMA-Net (ResNet-50)74.1Visual and Textual Prior Guided Mask Assemble for Few-Shot Segmentation and Beyond-
RePRI (ResNet-50)59.7Few-Shot Segmentation Without Meta-Learning: A Good Transductive Inference Is All You Need?
SegGPT (ViT)83.2SegGPT: Segmenting Everything In Context
MIANet (ResNet-50)68.72MIANet: Aggregating Unbiased Instance and General Information for Few-Shot Semantic Segmentation-
MSDNet (ResNet-101)64.7MSDNet: Multi-Scale Decoder for Few-Shot Semantic Segmentation via Transformer-Guided Prototyping
0 of 105 row(s) selected.