AI Paper Atlas

Research map / Benchmarks / VOC21

VOC21 leaderboard

12 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on VOC21. Best mIoU: 79.6 (ReME, ICCV 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.

VOC21 · mIoU

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

#mIoU ↑MethodPaperTopic
179.6ReME
CLIP ViT-B + DINOv2 ViT-L
ReME: A Data-Centric Framework for Training-Free Open-Vocabulary Segmentation
ICCV 2025 · Table 1
Open-set and few-shot detection
275.32PH-Reg
CLIP
Vision Transformers with Self-Distilled Registers
NeurIPS 2025 · Table: Linear Probe Based Eval
Attention and transformers
374.8CorrCLIP
CLIP ViT-B/16
CorrCLIP: Reconstructing Patch Correlations in CLIP for Open-Vocabulary Semantic Segmentation
ICCV 2025 · Table 1
Open-set and few-shot detection
472.8S2C2Seg + ProxyCLIP
CLIP ViT-B/16
S2C2Seg: Semantic-Spatial Consistency and Category Optimization for Open-Vocabulary Segmentation
CVPR 2026 · Table 1
Open-set and few-shot detection
565.8SCLIP + FreeCP
CLIP ViT-B/16
Training-Free Class Purification for Open-Vocabulary Semantic Segmentation
ICCV 2025 · Table 2
Open-set and few-shot detection
664.8RF-CLIP
CLIP ViT-B/16
Target Refocusing via Attention Redistribution for Open-Vocabulary Semantic Segmentation: An Explainability Perspective
AAAI 2026 · Table 2
Open-set and few-shot detection
764.2DIH-CLIP
CLIP ViT-B/16
DIH-CLIP: Unleashing the Diversity of Multi-Head Self-Attention for Training-Free Open-Vocabulary Semantic Segmentation
ICCV 2025 · Table 1
CNN features
863.9SFP
CLIP ViT-B/16
Feature Purification Matters: Suppressing Outlier Propagation for Training-Free Open-Vocabulary Semantic Segmentation
ICCV 2025 · Table 1
Open-set and few-shot detection
962.1LaVG
DINO ViT-B/8 + SCLIP ViT-B/16
In Defense of Lazy Visual Grounding for Open-Vocabulary Semantic Segmentation
ECCV 2024 · Table 2
Open-set and few-shot detection
1061.1NACLIP + ResCLIP
ViT-B/16
ResCLIP: Residual Attention for Training-free Dense Vision-language Inference
CVPR 2025 · Table 2
Human and embodied modeling
2 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.