HyperAIHyperAI

Command Palette

Search for a command to run...

Console

Semantic Segmentation On Cityscapes Val

Metrics

mIoU

Results

Performance results of various models on this benchmark

Paper Title
SERNet-Former87.35SERNet-Former: Semantic Segmentation by Efficient Residual Network with Attention-Boosting Gates and Attention-Fusion Networks
MetaPrompt-SD87.1Harnessing Diffusion Models for Visual Perception with Meta Prompts
InternImage-H87InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions
HRNetV2-OCR+PSA86.93Polarized Self-Attention: Towards High-quality Pixel-wise Regression
InternImage-XL86.4InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions
HRNet-OCR86.3Hierarchical Multi-Scale Attention for Semantic Segmentation
Depth Anything86.2Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data
ViT-Adapter-L85.8Vision Transformer Adapter for Dense Predictions
OneFormer (ConvNeXt-XL, Mapillary, multi-scale)85.8OneFormer: One Transformer to Rule Universal Image Segmentation
SeMask (SeMask Swin-L Mask2Former)84.98SeMask: Semantically Masked Transformers for Semantic Segmentation
Soft Labells (HRnet)84.8Soft labelling for semantic segmentation: Bringing coherence to label down-sampling
Sequential Ensemble (MiT-B5 + HRNet)84.8Sequential Ensembling for Semantic Segmentation
OneFormer (ConvNeXt-XL, multi-scale)84.6OneFormer: One Transformer to Rule Universal Image Segmentation
DiNAT-L (Mask2Former)84.5Dilated Neighborhood Attention Transformer
VPNeXt84.4VPNeXt -- Rethinking Dense Decoding for Plain Vision Transformer
OneFormer (Swin-L, multi-scale)84.4OneFormer: One Transformer to Rule Universal Image Segmentation
Mask2Former (Swin-L)84.3Masked-attention Mask Transformer for Universal Image Segmentation
VOLO-D4 (MS, ImageNet1k pretrain)84.3VOLO: Vision Outlooker for Visual Recognition
SegFormer (MiT-B5, Mapillary)84.0SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers
DDP (ConvNeXt-L, step-3)83.9DDP: Diffusion Model for Dense Visual Prediction
0 of 97 row(s) selected.