AI Paper Atlas

Research map / Benchmarks / PASCAL Context

PASCAL Context leaderboard

88 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on PASCAL Context. Best mIoU: 63.6 (Cat-Seg+R-SC-V, ICLR 2025).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

PASCAL Context · mIoU

54 papers, each with its best reported number for its own method; higher is better.

#mIoU ↑MethodPaperTopic
163.6Cat-Seg+R-SC-V
CLIP ViT-L/14
Refining CLIP's Spatial Awareness: A Visual-Centric Perspective
ICLR 2025 · Table: Results on open-vocabul
CLIP and zero-shot
262.3Cat-Seg+CLIPSelf
OpenAI ViT-L/14
CLIPSelf: Vision Transformer Distills Itself for Open-Vocabulary Dense Prediction
ICLR 2024 · Table: Results on open-vocabul
Open-set and few-shot detection
359.7NeCo
DINOv2R ViT-S/14
Near, far: Patch-ordering enhances vision foundation models' scene understanding
ICLR 2025 · Table: Evaluation of Full Fine
Open-set and few-shot detection
458.97UperNet + CAR
Swin-Large
CAR: Class-Aware Regularizations for Semantic Segmentation
ECCV 2022 · Table 2
CNN features
558.5HRFormer-B + OCRHRFormer: High-Resolution Vision Transformer for Dense Predict
NeurIPS 2021 · Table 5
Attention and transformers
658.2ProLab (Properties)
ViT-Adapter DeiT-B
A Semantic Space is Worth 256 Language Descriptions: Make Stronger Segmentation Models with Descriptive Properties
ECCV 2024 · Table 1
Open-set and few-shot detection
758.2SETR + DToP@Finetune (p0=0.98)
ViT-Base
Dynamic Token Pruning in Plain Vision Transformers for Semantic Segmentation
ICCV 2023 · Table 7
CNN features
857.2OCR + Ours
HRNetV2-W48
Cross-Image Pixel Contrasting for Semantic Segmentation
TPAMI 2024 · Table 5
CNN features
957.2OCR + Ours
HRNetV2-W48
Exploring Cross-Image Pixel Contrast for Semantic Segmentation
ICCV 2021 · Table 5
CNN features
1057.15Falcon
SSD-1B
Falcon: Fast Proximal Linearization of Normalized Cuts for Unsupervised Image Segmentation
ICLR 2026 · Table 1
Open-set and few-shot detection
44 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

PASCAL Context · mIoU (val)

35 papers, each with its best reported number for its own method; higher is better.

#mIoU ↑MethodPaperTopic
164.51UPerNet+IDRNet
Swin-Large
IDRNet: Intervention-Driven Relation Network for Semantic Segmentation
NeurIPS 2023 · Table SOTA
CNN features
263.3TSG
Swin-Large
Transformer Scale Gate for Semantic Segmentation
CVPR 2023 · Table 1
CNN features
355.83SETR-MLA (80k, MS)
T-Large
Rethinking Semantic Segmentation From a Sequence-to-Sequence Perspective With Transformers
CVPR 2021 · Table: State-of-the-art compar
CNN features
455.8STLNet
ResNet-101
Learning Statistical Texture for Semantic Segmentation
CVPR 2021 · Table: Results comparison on P
CNN features
555.3EfficientFCN (HGD)
ResNet101
A Holistically-Guided Decoder for Deep Representation Learning With Applications to Semantic Segmentation and Object Detection
TPAMI 2021 · Table 1
CNN features
655.3EfficientFCN
ResNet101
EfficientFCN: Holistically-guided Decoding for Semantic Segmentation
ECCV 2020 · Table 5
CNN features
755.2HamNet
ResNet-101
Is Attention Better Than Matrix Decomposition?
ICLR 2021 · Table: Results on the PASCAL-C
Attention and transformers
854.9RANet
ResNet-101
RANet: Region Attention Network for Semantic Segmentation
NeurIPS 2020 · Table 6
CNN features
954CFNet
ResNet-101
Co-Occurrent Features in Semantic Segmentation
CVPR 2019 · Table 4
CNN features
1053.9CPNet101
ResNet-101
Context Prior for Scene Segmentation
CVPR 2020 · Table: Quantitative evaluation
CNN features
25 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.