For adding your transformer-based object detector results into the tables below, please send us an email including the values for each column and a copy of the paper showing your results.
Email: [email protected]
Taxonomy of small object detection using transformers and popular object detection methods assigned to each category.
Detection performance for small-scale objects on MS COCO image dataset (eval). DC5: Dialated C5 stage, MS: Multi-scale network, IBR: Iterative bounding box refinement, TS: Two-stage detection, DCN: Deformable convnets, TTA: Test time augmentation, BD: Pre-trained on BigDetection dataset, IN: Pre-trained on ImageNet, OB: Pre-trained on Object-365.
| Model | Backbone | GFLOPS/FPS | #params | Epochs | URL | |
|---|---|---|---|---|---|---|
| Faster RCNN-DC5~(NeurIPS2015) | ResNet50 | 320/16 | 166M | 21.4 | 37 | https://github.com/trzy/FasterRCNN |
| Faster RCNN-FPN~(NeurIPS2015) | ResNet50 | 180/26 | 42M | 24.2 | 37 | https://github.com/trzy/FasterRCNN |
| Faster RCNN-FPN~(NeurIPS2015) | ResNet101 | 246/20 | 60M | 25.2 | -- | https://github.com/trzy/FasterRCNN |
| RepPoints v2-DCN-MS~(NeurIPS2020) | ResNeXt101 | --/-- | -- | 34.5* | 24 | https://github.com/Scalsol/RepPointsV2 |
| FCOS~(ICCV2019) | ResNet50 | 177/17 | -- | 26.2 | 36 | https://github.com/tianzhi0549/FCOS |
| CBNet V2-DCN~(TIP2022) | Res2Net101 | --/-- | 107M | 35.7* | 20 | https://github.com/VDIGPKU/CBNetV2 |
| CBNet V2-DCN(Cascade RCNN)~(TIP2022) | Res2Net101 | --/-- | 146M | 37.4* | 32 | https://github.com/VDIGPKU/CBNetV2 |
| DETR~(ECCV2020) | ResNet50 | 86/28 | 41M | 20.5 | 500 | https://github.com/facebookresearch/detr |
| DETR-DC5~(ECCV2020) | ResNet50 | 187/12 | 41M | 22.5 | 500 | https://github.com/facebookresearch/detr |
| DETR~(ECCV2020) | ResNet101 | 52/20 | 60M | 21.9 | -- | https://github.com/facebookresearch/detr |
| DETR-DC5~(ECCV2020) | ResNet101 | 253/10 | 60M | 23.7 | -- | https://github.com/facebookresearch/detr |
| ViT-FRCNN~(arXiv2020) | -- | --/-- | -- | 17.8 | -- | -- |
| RelationNet++~(NeurIPS2020) | ResNeXt101 | --/-- | -- | 32.8* | -- | https://github.com/microsoft/RelationNet2 |
| RelationNet++-MS~(NeurIPS2020) | ResNeXt101 | --/-- | -- | 35.8* | -- | https://github.com/microsoft/RelationNet2 |
| Deformable DETR~(ICLR2021) | ResNet50 | 173/19 | 40M | 26.4 | 50 | https://github.com/fundamentalvision/Deformable-DETR |
| Deformable DETR-IBR~(ICLR2021) | ResNet50 | 173/19 | 40M | 26.8 | 50 | https://github.com/fundamentalvision/Deformable-DETR |
| Deformable DETR-TS~(ICLR2021) | ResNet50 | 173/19 | 40M | 28.8 | 50 | https://github.com/fundamentalvision/Deformable-DETR |
| Deformable DETR-TS-IBR-DCN~(ICLR2021) | ResNeXt101 | --/-- | -- | 34.4* | -- | https://github.com/fundamentalvision/Deformable-DETR |
| Dynamic DETR~(ICCV2021) | ResNet50 | --/-- | -- | 28.6* | -- | -- |
| Dynamic DETR-DCN~(ICCV2021) | ResNeXt101 | --/-- | -- | 30.3* | -- | -- |
| TSP-FCOS~(ICCV2021) | ResNet101 | 255/12 | -- | 27.7 | 36 | https://github.com/Edward-Sun/TSP-Detection |
| TSP-RCNN~(ICCV2021) | ResNet101 | 254/9 | -- | 29.9 | 96 | https://github.com/Edward-Sun/TSP-Detection |
| Mask R-CNN~(ICCV2021) | Conformer-S/16 | 457/-- | 56.9M | 28.7 | 12 | https://github.com/pengzhiliang/Conformer |
| Conditional DETR-DC5~(ICCV2021) | ResNet101 | 262/-- | 63M | 27.2 | 108 | https://github.com/Atten4Vis/ConditionalDETR |
| SOF-DETR~(2022JVCIR) | ResNet50 | --/-- | -- | 21.7 | -- | https://github.com/shikha-gist/SOF-DETR/ |
| DETR++~(arXiv2022) | ResNet50 | --/-- | -- | 22.1 | -- | -- |
| TOLO-MS~(NCA2022) | -- | --/57 | -- | 24.1 | -- | -- |
| Anchor DETR-DC5~(AAAI2022) | ResNet101 | --/-- | -- | 25.8 | 50 | https://github.com/megvii-research/AnchorDETR |
| DESTR-DC5~(CVPR2022) | ResNet101 | 299/-- | 88M | 28.2 | 50 | -- |
| Conditional DETR v2-DC5~(arXiv2022) | ResNet101 | 228/-- | 65M | 26.3 | 50 | -- |
| Conditional DETR v2~(arXiv2022) | Hourglass48 | 521/-- | 90M | 32.1 | 50 | -- |
| FP-DETR-IN~(ICLR2022) | -- | --/-- | 36M | 26.5 | 50 | https://github.com/encounter1997/FP-DETR |
| DAB-DETR-DC5~(arXiv2022) | ResNet101 | 296/-- | 63M | 28.1 | 50 | https://github.com/IDEA-Research/DAB-DETR |
| Ghostformer-MS~(Sensors2022) | GhostNet | --/-- | -- | 29.2 | 100 | -- |
| CF-DETR-DCN-TTA~(AAAI2022) | ResNeXt101 | --/-- | -- | 35.1* | -- | -- |
| CBNet V2-TTA~(CVPR2022) | Swin Transformer-base | --/-- | -- | 41.7 | -- | https://github.com/amazon-science/bigdetection |
| CBNet V2-TTA-BD~(CVPR2022) | Swin Transformer-base | --/-- | -- | 42.2 | -- | https://github.com/amazon-science/bigdetection |
| DETA~(arXiv2022) | ResNet50 | --/13 | 48M | 34.3 | 24 | https://github.com/jozhang97/DETA |
| DINO~(arXiv2022) | ResNet50 | 860/10 | 47M | 32.3 | 12 | https://github.com/IDEA-Research/DINO |
| CO-DINO Deformable DETR-MS-IN~(arXiv2022) | Swin Transformer-large | --/-- | -- | 43.7 | 36 | https://github.com/Sense-X/Co-DETR |
| HYNETER~(ICASSP2023) | Hyneter-Max | --/-- | 247M | 29.8* | -- | -- |
| DeoT~(JRTIP2023) | ResNet101 | 217/14 | 58M | 31.4 | 34 | -- |
| ConformerDet-MS~(TPAMI2023) | Conformer-B | --/-- | 147M | 35.3 | 36 | https://github.com/pengzhiliang/Conformer |
| YOLOS~(NeurIPS2021) | DeiT-base | --/3.9 | 100M | 19.5 | 150 | https://github.com/hustvl/YOLOS |
| DETR(ViT)~(arXiv2021) | Swin Transformer-base | --/9.7 | 100M | 18.3 | 50 | https://github.com/naver-ai/vidt |
| Deformable DETR(ViT)~(arXiv2021) | Swin Transformer-base | --/4.8 | 100M | 34.5 | 50 | https://github.com/naver-ai/vidt |
| ViDT~(arXiv2022) | Swin Transformer-base | --/9 | 100M | 30.6 | 50 | https://github.com/naver-ai/vidt/tree/main |
| DFFT~(ECCV2022) | DOT-medium | 67/-- | -- | 25.5 | 36 | https://github.com/PeixianChen/DFFT |
| CenterNet++-MS~(arXiv2022) | Swin Transformer-large | --/-- | -- | 38.7* | -- | https://github.com/Duankaiwen/PyCenterNet |
| DETA-OB~(arXiv2022) | Swin Transformer-large | --/4.2 | -- | 46.1* | 24 | https://github.com/jozhang97/DETA |
| Group DETR v2-MS-IN-OB~(arXiv2022) | ViT-Huge | --/-- | 629M | 48.4* | -- | -- |
| Best Results | NA | DETR/TOLO | FP-DETR | Group DETR v2 | DINO | NA |
Detection performance for objects on DOTA image dataset. MS: Multi-scale network, FT: Fine-tuned, FPN: Feature pyramid network, IN: Pre-trained on ImageNet.
| Model | Backbone | FPS | #params | mAP | Epochs | URL |
|---|---|---|---|---|---|---|
| Rotated Faster RCNN-MS~(NeurIPS2015) | ResNet101 | -- | 64M | 67.71 | 50 | https://github.com/open-mmlab/mmrotate/tree/main/configs/rotated_faster_rcnn |
| SSD~(ECCV2016) | -- | -- | -- | 56.1 | -- | https://github.com/pierluigiferrari/ssd_keras |
| RetinaNet-MS~(ICCV2017) | ResNet101 | -- | 59M | 66.53 | 50 | https://github.com/DetectionTeamUCAS/RetinaNet_Tensorflow |
| ROI-Transformer-MS-IN~(CVPR2019) | ResNet50 | -- | -- | 80.06 | 12 | https://github.com/open-mmlab/mmrotate/blob/main/configs/roi_trans/README.md |
| Yolov5~(2020) | -- | 95 | -- | 64.5 | -- | https://github.com/ultralytics/yolov5 |
| ReDet-MS-FPN~(CVPR2021) | ResNet50 | -- | -- | 80.1 | -- | https://github.com/csuhan/ReDet |
| O2DETR-MS~(arXiv2021) | ResNet101 | -- | 63M | 70.02 | 50 | -- |
| O2DETR-MS-FT~(arXiv2021) | ResNet101 | -- | -- | 76.23 | 62 | -- |
| O2DETR-MS-FPN-FT~(arXiv2021) | ResNet50 | -- | -- | 79.66 | -- | -- |
| SPH-Yolov5~(RS2022) | Swin Transformer-base | 51 | -- | 71.6 | 150 | -- |
| AO2-DETR-MS~(TCSVT2022) | ResNet50 | -- | -- | 79.22 | -- | https://github.com/Ixiaohuihuihui/AO2-DETR |
| MDCT~(RS2023) | -- | -- | -- | 75.7 | -- | -- |
| ReDet-MS-IN~(arXiv2023) | ViTDet, ViT-B | -- | -- | 80.89 | 12 | https://github.com/csuhan/ReDet |
| Best Results | NA | Yolov5 | RetinaNet | ReDet-MS-IN | ReDet-MS-IN | NA |
Detection performance for DeepLesion CT image dataset.
| Model | Accuracy | |
|---|---|---|
| Faster RCNN~(NeurIPS2015) | 83.3 | 83.3 |
| Yolov5 | 85.2 | 88.2 |
| DETR~(ECCV2020) | 86.7 | 87.8 |
| Swin Transformer | 82.9 | 81.2 |
| MS Transformer~(CIN2022) | 90.3 | 89.6 |
| Best Results | MS Transformer | MS Transformer |
Detection performance for AMWW image dataset.
| Model | Backbone | ||
|---|---|---|---|
| Faster RCNN~(NeurIPS2015) | ResNet50 | 70.7 | 26.83 |
| Cascade RCNN~(CVPR2018) | ResNet50 | 74.7 | 27.8 |
| TridentNet~(ICCV2019) | ResNet50 | 77.3 | 29.2 |
| Dynamic RCNN~(ECCV2020) | ResNet50 | 76.3 | 27.6 |
| Yolov5 | ResNet50 | 76.67 | 28.48 |
| MATR~(TCSVT2022) | ResNet50 | 82.16 | 33.42 |
| Best Results | NA | MATR | MATR |
Detection performance for URPC2018 dataset.
| Model | #params | ||
|---|---|---|---|
| Faster RCNN~(NeurIPS2015) | 33.6M | 16.4 | -- |
| Cascade RCNN~(CVPR2018) | 68.9M | 16 | -- |
| Dynamic RCNN~(ECCV2020) | 41.5M | 13.3 | -- |
| Yolov3 | 61.5M | 19.4 | -- |
| RoIMix~(ICASSP2020) | -- | -- | 74.92 |
| HTDet~(RS2023) | 7.7M | 22.8 | -- |
| Best Results | HTDet | HTDet | RoIMix |
Detection performance for ImageNet VID dataset for small objects. PT: Pre-trained on MS COCO.
| Model | Backbone | |
|---|---|---|
| Faster RCNN~(NeurIPS2015)+SELSA | ResNet50 | 8.5 |
| Deformable-DETR-PT | ResNet50 | 10.5 |
| Deformable-DETR+TransVOD-PT | ResNet50 | 11 |
| DAB-DETR+FAQ-PT | ResNet50 | 12 |
| Deformable-DETR+FAQ-PT | ResNet50 | 13.2 |
| Best Results | NA | Deformable DET+FAQ |
Detection results on a sample image when zoomed in. First row from the left: Input image, SSD, Faster RCNN, DETR. Second row from the left: ViDT, DETA-OB, DINO, CBNetv2.
If you found this page helpful, please cite the following survey papers:
@article{rekavandi2023transformers,
title={Transformers in Small Object Detection: A Benchmark and Survey of State-of-the-Art},
author={Rekavandi Miri, Aref and Rashidi, Shima and Boussaid, Farid and Hoefs, Stephen and Akbas, Emre and Bennamoun, Mohammed},
journal={arXiv preprint arXiv:2309.04902},
year={2023}
}
@article{rekavandi2022guide,
title={A Guide to Image and Video based Small Object Detection using Deep Learning: Case Study of Maritime Surveillance},
author={Rekavandi Miri, Aref and Xu, Lian and Boussaid, Farid and Seghouane, Abd-Krim and Hoefs, Stephen and Bennamoun, Mohammed},
journal={arXiv preprint arXiv:2207.12926},
year={2022}
}