| Rank | Model |
mask AP | AP50 | AP75 | APS | APM | APL |
Extra Training Data |
Paper | Code | Year |
| 1 |
Co-DETR |
57.1 | 80.2 | 63.4 | 41.6 | 60.1 | 72.0 |
✓ |
DETRs with Collaborative Hybrid Assignments Training
|
open-mmlab/mmdetection · siyuanliii/masa · sense-x/co-detr
· +3 |
2022 |
| 2 |
CBNetV2 (EVA02, single-scale) |
56.1 | 80.3 | 62.1 | 39.7 | 59.3 | 70.9 |
✓ |
CBNet: A Composite Backbone Network Architecture for Object Detection
|
PaddlePaddle/PaddleDetection · shinya7y/UniverseNet · VDIGPKU/CBNetV2
· +1 |
2021 |
| 3 |
EVA |
55.5 | 80.0 | – | 36.3 | 58.0 | 72.4 |
✓ |
EVA: Exploring the Limits of Masked Visual Representation Learning at Scale
|
rwightman/pytorch-image-models · open-mmlab/mmselfsup · baaivision/eva
· +3 |
2022 |
| 4 |
FD-SwinV2-G |
55.4 | – | – | – | – | – |
✓ |
Contrastive Learning Rivals Masked Image Modeling in Fine-tuning via Feature Distillation
|
SwinTransformer/Feature-Distillation |
2022 |
| 5 |
Mask Frozen-DETR |
55.3 | 79.3 | 61.4 | 37.8 | 58.4 | 70.4 |
✓ |
Mask Frozen-DETR: High Quality Instance Segmentation with One GPU
|
|
2023 |
| 6 |
BEiT-3 |
54.8 | – | – | – | – | – |
|
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks
|
microsoft/unilm · lyan62/data-curation |
2022 |
| 7 |
MasK DINO (SwinL, multi-scale) |
54.7 | – | – | – | – | – |
✓ |
Mask DINO: Towards A Unified Transformer-based Framework for Object Detection and Segmentation
|
PaddlePaddle/PaddleDetection · IDEACVR/DINO · idea-research/maskdino
· +7 |
2022 |
| 8 |
ViT-Adapter-L (HTC++, BEiTv2, O365, multi-scale) |
54.5 | – | – | – | – | – |
✓ |
Vision Transformer Adapter for Dense Predictions
|
czczup/vit-adapter · chenller/mmseg-extension |
2022 |
| 8 |
GLEE-Pro |
54.5 | – | – | – | – | – |
✓ |
General Object Foundation Model for Images and Videos at Scale
|
FoundationVision/GLEE |
2023 |
| 10 |
SwinV2-G (HTC++) |
54.4 | – | – | – | – | – |
✓ |
Swin Transformer V2: Scaling Up Capacity and Resolution
|
rwightman/pytorch-image-models · microsoft/Swin-Transformer · PaddlePaddle/PaddleDetection
· +20 |
2021 |
| 11 |
GLEE-Plus |
53.3 | – | – | – | – | – |
✓ |
General Object Foundation Model for Images and Videos at Scale
|
FoundationVision/GLEE |
2023 |
| 12 |
Soft Teacher + Swin-L (HTC++, multi-scale) |
53.0 | – | – | – | – | – |
✓ |
End-to-End Semi-Supervised Object Detection with Soft Teacher
|
microsoft/SoftTeacher · amazon-science/bigdetection · amazon-research/bigdetection
· +5 |
2021 |
| 12 |
ViT-Adapter-L (HTC++, BEiTv2 pretrain, multi-scale) |
53.0 | – | – | – | – | – |
|
Vision Transformer Adapter for Dense Predictions
|
czczup/vit-adapter · chenller/mmseg-extension |
2022 |
| 14 |
Mask DINO (SwinL, single -scale) |
52.8 | – | – | – | – | – |
|
Mask DINO: Towards A Unified Transformer-based Framework for Object Detection and Segmentation
|
PaddlePaddle/PaddleDetection · IDEACVR/DINO · idea-research/maskdino
· +7 |
2022 |
| 15 |
ViT-Adapter-L (HTC++, BEiT pretrain, multi-scale) |
52.5 | – | – | – | – | – |
|
Vision Transformer Adapter for Dense Predictions
|
czczup/vit-adapter · chenller/mmseg-extension |
2022 |
| 16 |
CBNetV2 (Dual-Swin-L HTC, multi-scale) |
52.3 | – | – | – | – | – |
|
CBNet: A Composite Backbone Network Architecture for Object Detection
|
PaddlePaddle/PaddleDetection · shinya7y/UniverseNet · VDIGPKU/CBNetV2
· +1 |
2021 |
| 17 |
UNINEXT-H |
51.8 | 76.2 | 56.7 | 33.3 | 55.9 | 67.5 |
✓ |
Universal Instance Perception as Object Discovery and Retrieval
|
MasterBin-IIAU/UNINEXT |
2023 |
| 18 |
CBNetV2 (Dual-Swin-L HTC, single-scale) |
51.6 | – | – | – | – | – |
|
CBNet: A Composite Backbone Network Architecture for Object Detection
|
PaddlePaddle/PaddleDetection · shinya7y/UniverseNet · VDIGPKU/CBNetV2
· +1 |
2021 |
| 19 |
Focal-L (HTC++, multi-scale) |
51.3 | 75.4 | 56.5 | 35.6 | – | 64.2 |
|
Focal Self-attention for Local-Global Interactions in Vision Transformers
|
BR-IDL/PaddleViT · microsoft/Focal-Transformer · microsoft/esvit |
2021 |
| 20 |
Swin-L (HTC++, multi scale) |
51.1 | – | – | – | – | – |
|
Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
|
huggingface/transformers · rwightman/pytorch-image-models · open-mmlab/mmdetection
· +77 |
2021 |