HyperAI超神经

Semantic Segmentation On Cityscapes

评估指标

Mean IoU (class)

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 (class)
Paper TitleRepository
ESANet-R34-NBt1D80.09%Efficient RGB-D Semantic Segmentation for Indoor Scene Analysis-
ESNet70.7%ESNet: An Efficient Symmetric Network for Real-time Semantic Segmentation
ResNeSt200 (Mapillary)83.3%ResNeSt: Split-Attention Networks
DUC-HDC (ResNet-101)77.6%Understanding Convolution for Semantic Segmentation
LightSeg-DarkNet1970.75%LiteSeg: A Novel Lightweight ConvNet for Semantic Segmentation
AdapNet++81.24%Self-Supervised Model Adaptation for Multimodal Semantic Segmentation
HRNetV2 + OCR (w/ ASP)83.7%Segmentation Transformer: Object-Contextual Representations for Semantic Segmentation
ShelfNet-3479.0%ShelfNet for Fast Semantic Segmentation
DeepLabv3 (ResNet-101, coarse)81.3%Rethinking Atrous Convolution for Semantic Image Segmentation
InternImage-H86.1%InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions
SqueezeNAS (LAT Large)72.5%SqueezeNAS: Fast neural architecture search for faster semantic segmentation
OCNet81.7%OCNet: Object Context Network for Scene Parsing
MRFM(coarse)83.0%Multi Receptive Field Network for Semantic Segmentation-
LiteSeg-MobileNet67.81%LiteSeg: A Novel Lightweight ConvNet for Semantic Segmentation
HRNetV2 (train+val)81.6%Deep High-Resolution Representation Learning for Visual Recognition
ESPNetv266.2%ESPNetv2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network
LightSeg-MobileNet67.81%LiteSeg: A Novel Lightweight ConvNet for Semantic Segmentation
DPN66.8%Semantic Image Segmentation via Deep Parsing Network
SPNet (ResNet-101)82.0%Strip Pooling: Rethinking Spatial Pooling for Scene Parsing
HRNet (HRNetV2-W48)81.6%High-Resolution Representations for Labeling Pixels and Regions
0 of 105 row(s) selected.