HyperAI超神经
首页资讯最新论文教程数据集百科SOTALLM 模型天梯GPU 天梯顶会
全站搜索
关于
HyperAI超神经

Semantic Segmentation On Pascal Voc 2012 Val

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

模型名称
mIoU
Paper TitleRepository
DeepLabv3-JFT82.7%Rethinking Atrous Convolution for Semantic Image Segmentation
RRM66.3Reliability Does Matter: An End-to-End Weakly Supervised Semantic Segmentation Approach
FastDenseNas-arch177.1%Fast Neural Architecture Search of Compact Semantic Segmentation Models via Auxiliary Cells
SID-Simple Does It: Weakly Supervised Instance and Semantic Segmentation-
SpineNet-S143 (single-scale test)85.64%Dilated SpineNet for Semantic Segmentation-
FastDenseNas-arch277.3%Fast Neural Architecture Search of Compact Semantic Segmentation Models via Auxiliary Cells
PRM53.4%Weakly Supervised Instance Segmentation using Class Peak Response
Auto-DeepLab-L82.04%Auto-DeepLab: Hierarchical Neural Architecture Search for Semantic Image Segmentation
HyperSeg-L80.61%HyperSeg: Patch-wise Hypernetwork for Real-time Semantic Segmentation
SSDD64.9Self-Supervised Difference Detection for Weakly-Supervised Semantic Segmentation
PSA w/ EADER DeepLab (Xception-65)62.8%Find it if You Can: End-to-End Adversarial Erasing for Weakly-Supervised Semantic Segmentation
FastDenseNas-arch078.0%Fast Neural Architecture Search of Compact Semantic Segmentation Models via Auxiliary Cells
ReLICv277.9%Pushing the limits of self-supervised ResNets: Can we outperform supervised learning without labels on ImageNet?
Eff-B7 NAS-FPN (Copy-Paste pre-training, single-scale))86.6%Simple Copy-Paste is a Strong Data Augmentation Method for Instance Segmentation
BYOL75.7%Pushing the limits of self-supervised ResNets: Can we outperform supervised learning without labels on ImageNet?
SIW65%Scaling up Multi-domain Semantic Segmentation with Sentence Embeddings-
G255.7%Exploiting saliency for object segmentation from image level labels-
DetCon77.3%Pushing the limits of self-supervised ResNets: Can we outperform supervised learning without labels on ImageNet?
DeepLabv3 (ImageNet+300M)76.5%Revisiting Unreasonable Effectiveness of Data in Deep Learning Era
TADP87.11%Text-image Alignment for Diffusion-based Perception
0 of 29 row(s) selected.