Browse › Computer Vision › Object Detection › COCO test-dev
Object Detection archive 2025-07-28
COCO test-dev Benchmark (Object Detection)
Over time archive 2025-07-28
The chart needs JavaScript; the table below carries every value.
Direction inferred from the metric name, not from the archive: box mAP (higher is better), Params (M) (lower is better). Not inferred (points only, no best-so-far line): AP50, AP75, APS, APM, APL, Hardware Burden, Operations per network pass, GFLOPs. Points are placed at the row's paper date; 225 of 225 rows carry one.
Results archive 2025-07-28
Archive rows end at the archive snapshot, 2025-07-28: no result published after that date is in this table. Rank is the archive's row order at that snapshot; not re-ranked here. Metric values are the archive's strings. Column headers sort the table in your browser; each row keeps its archive rank.
| Paper | Code | Ran Syntology | Report | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Co-DETR | 66.0 | 304 | – | Paper | Code | 2022 | 0 of 5 ran · 5 unverified | report | ||||||||
| 2 | InternImage-H (M3I Pre-training) | 65.5 | 2180 | – | Paper | Code | 2022 | 2 of 4 ran · 2 unverified | report | ||||||||
| 3 | M3I Pre-training (InternImage-H) | 65.4 | – | Paper | Code | 2022 | linked, not harvested | report | |||||||||
| 4 | MoCaE | 65.1 | – | Paper | Code | 2023 | 1 of 1 ran · 0 unverified | report | |||||||||
| 5 | Focal-Stable-DINO (Focal-Huge, no TTA) | 64.8 | 81.7 | 71.5 | 48.6 | 67.6 | 78 | 689 | – | Paper | Code | 2023 | linked, not harvested | report | |||
| 6 | Co-DETR (Swin-L) | 64.8 | 218 | – | Paper | Code | 2022 | 0 of 5 ran · 5 unverified | report | ||||||||
| 7 | EVA | 64.7 | 81.9 | 71.7 | 48.5 | 67.7 | 77.9 | – | Paper | Code | 2022 | 1 of 3 ran · 2 unverified | report | ||||
| 8 | Group DETR v2 | 64.5 | 81.8 | 71.1 | 48.4 | 67.2 | 77.1 | – | Paper | – | 2022 | no code linked | report | ||||
| 9 | FocalNet-H (DINO) | 64.4 | – | Paper | Code | 2022 | 0 of 2 ran · 2 unverified | report | |||||||||
| 10 | InternImage-XL | 64.3 | 602 | – | Paper | Code | 2022 | 2 of 4 ran · 2 unverified | report | ||||||||
| 11 | FD-SwinV2-G | 64.2 | – | Paper | Code | 2022 | 2 of 8 ran · 6 unverified | report | |||||||||
| 12 | Plain-DETR (Swin-L) | 63.9 | 82.1 | 70.7 | 48.2 | 66.8 | 76.7 | 228 | – | Paper | Code | 2023 | linked, not harvested | report | |||
| 13 | RevCol-H(DINO) | 63.8 | – | Paper | Code | 2022 | 8 of 11 ran · 3 unverified | report | |||||||||
| 14 | BEiT-3 | 63.7 | – | Paper | Code | 2022 | linked, not harvested | report | |||||||||
| 15 | Relation-DETR (Focal-L) | 63.5 | 80.8 | 69.1 | 47.2 | 66.9 | 77.0 | 214 | – | Paper | Code | 2024 | 1 of 1 ran · 0 unverified | report | |||
| 16 | DETA (Swin-L) | 63.5 | 80.4 | 70.2 | 46.1 | 66.9 | 76.9 | – | Paper | Code | 2022 | linked, not harvested | report | ||||
| 17 | DINO (Swin-L,multi-scale, TTA) | 63.3 | – | Paper | Code | 2022 | 7 of 15 ran · 8 unverified | report | |||||||||
| 18 | SwinV2-G (HTC++) | 63.1 | 3000 | – | Paper | Code | 2021 | 3 of 30 ran · 27 unverified | report | ||||||||
| 19 | Grounding DINO | 63.0 | – | Paper | Code | 2023 | 2 of 5 ran · 3 unverified | report | |||||||||
| 20 | Florence-CoSwin-H | 62.4 | – | Paper | Code | 2021 | linked, not harvested | report | |||||||||
| 21 | GLIPv2 (CoSwin-H, multi-scale) | 62.4 | – | Paper | Code | 2022 | 2 of 2 ran · 0 unverified | report | |||||||||
| 22 | GLEE-Pro | 62.3 | – | Paper | Code | 2023 | 8 of 13 ran · 5 unverified | report | |||||||||
| 23 | GLIP (Swin-L, multi-scale) | 61.5 | 79.5 | 67.7 | 45.3 | 64.9 | 75.0 | – | Paper | Code | 2021 | 2 of 2 ran · 0 unverified | report | ||||
| 24 | Soft Teacher + Swin-L (HTC++, multi-scale) | 61.3 | – | Paper | Code | 2021 | linked, not harvested | report | |||||||||
| 25 | ViT-Adapter-L (HTC++, BEiTv2 pretrain, multi-scale) | 60.9 | – | Paper | Code | 2022 | linked, not harvested | report | |||||||||
| 26 | DyHead (Swin-L, multi scale, self-training) | 60.6 | 78.5 | 66.6 | 64.0 | 74.2 | – | Paper | Code | 2021 | 7 of 9 ran · 2 unverified | report | |||||
| 27 | GLEE-Plus | 60.6 | – | Paper | Code | 2023 | 8 of 13 ran · 5 unverified | report | |||||||||
| 28 | ViT-Adapter-L (HTC++, BEiT pretrain, multi-scale) | 60.4 | – | Paper | Code | 2022 | linked, not harvested | report | |||||||||
| 29 | GRiT (ViT-H, single-scale testing) | 60.4 | – | Paper | Code | 2022 | 5 of 8 ran · 3 unverified | report | |||||||||
| 30 | CBNetV2 (Dual-Swin-L HTC, multi-scale) | 60.1 | – | Paper | Code | 2021 | 2 of 3 ran · 1 unverified | report | |||||||||
| 31 | PIIP-H6B (DINO) | 60.0 | 79.0 | 65.4 | – | Paper | Code | 2024 | linked, not harvested | report | |||||||
| 32 | CBNetV2 (Dual-Swin-L HTC, single-scale) | 59.4 | – | Paper | Code | 2021 | 2 of 3 ran · 1 unverified | report | |||||||||
| 33 | Focal-L (DyHead, multi-scale) | 58.9 | – | Paper | Code | 2021 | 3 of 3 ran · 0 unverified | report | |||||||||
| 34 | DyHead (Swin-L, multi scale) | 58.7 | 77.1 | 64.5 | 62.0 | 72.8 | – | Paper | Code | 2021 | 7 of 9 ran · 2 unverified | report | |||||
| 35 | Swin-L (HTC++, multi scale) | 58.7 | – | Paper | Code | 2021 | 108 of 207 ran · 99 unverified | report | |||||||||
| 36 | Swin-L (HTC++, single scale) | 57.7 | – | Paper | Code | 2021 | 108 of 207 ran · 99 unverified | report | |||||||||
| 37 | Cascade Eff-B7 NAS-FPN (1280, self-training Copy Paste, single-scale) | 57.3 | – | Paper | Code | 2020 | 2 of 2 ran · 0 unverified | report | |||||||||
| 38 | PyCenterNet (Swin-L, multi-scale) | 57.1 | 73.7 | 62.4 | 38.7 | 59.2 | 71.3 | – | Paper | Code | 2022 | linked, not harvested | report | ||||
| 39 | dBOT ViT-L (CLIP) | 56.8 | – | Paper | Code | 2022 | 5 of 13 ran · 8 unverified | report | |||||||||
| 40 | YOLOv7-D6 (44 fps) | 56.6 | ✓ | Paper | Code | 2022 | 1 of 11 ran · 10 unverified | report | |||||||||
| 41 | SOLQ (Swin-L, single scale) | 56.5 | 74.6 | 60.5 | 37.6 | 60 | 70.6 | – | Paper | Code | 2021 | linked, not harvested | report | ||||
| 42 | CenterNet2 (Res2Net-101-DCN-BiFPN, self-training, 1560 single-scale) | 56.4 | 74.0 | 61.6 | 38.7 | 59.7 | 68.6 | – | Paper | Code | 2021 | linked, not harvested | report | ||||
| 43 | ISTR (ResNet50-FPN-3x, single-scale) | 56.4 | 27.8 | 48.7 | 59.9 | – | Paper | Code | 2021 | linked, not harvested | report | ||||||
| 44 | QueryInst (single-scale) | 56.1 | 75.9 | 61.9 | 37.4 | 58.9 | 70.3 | 17G | – | Paper | Code | 2021 | 1 of 3 ran · 2 unverified | report | |||
| 45 | dBOT ViT-L | 56.1 | – | Paper | Code | 2022 | 5 of 13 ran · 8 unverified | report | |||||||||
| 46 | YOLOv7-E6 (56 fps) | 56 | – | Paper | Code | 2022 | 1 of 11 ran · 10 unverified | report | |||||||||
| 47 | YOLOv4-P7 with TTA | 55.8 | 73.2 | 61.2 | – | Paper | Code | 2020 | 1 of 1 ran · 0 unverified | report | |||||||
| 48 | DetectoRS (ResNeXt-101-64x4d, multi-scale) | 55.7 | 74.2 | 61.1 | 37.7 | 58.4 | 68.1 | – | Paper | Code | 2020 | 0 of 4 ran · 4 unverified | report | ||||
| 49 | YOLOR-D6 (1280, single-scale, 30 fps) | 55.4 | 73.3 | 60.6 | – | Paper | Code | 2021 | linked, not harvested | report | |||||||
| 50 | YOLOv4-P6 with TTA | 54.9 | 72.6 | 60.2 | – | Paper | Code | 2020 | 1 of 1 ran · 0 unverified | report | |||||||
| 51 | YOLOv7-W6 (84 fps) | 54.9 | – | Paper | Code | 2022 | 1 of 11 ran · 10 unverified | report | |||||||||
| 52 | Cascade Eff-B7 NAS-FPN (1280) | 54.8 | – | Paper | Code | 2020 | 2 of 2 ran · 0 unverified | report | |||||||||
| 53 | DetectoRS (ResNeXt-101-32x4d, multi-scale) | 54.7 | 73.5 | 60.1 | 37.4 | 57.3 | 66.4 | – | Paper | Code | 2020 | 0 of 4 ran · 4 unverified | report | ||||
| 54 | GLEE-Lite | 54.7 | – | Paper | Code | 2023 | 8 of 13 ran · 5 unverified | report | |||||||||
| 55 | YOLOv4-P6 CSP-P6 (single-scale, 32 fps) | 54.3 | 72.3 | 59.5 | 36.6 | 58.2 | 65.5 | – | Paper | Code | 2020 | 1 of 1 ran · 0 unverified | report | ||||
| 56 | SpineNet-190 (1280, with Self-training on OpenImages, single-scale) | 54.3 | – | Paper | Code | 2020 | linked, not harvested | report | |||||||||
| 57 | UniverseNet-20.08d (Res2Net-101, DCN, multi-scale) | 54.1 | 71.6 | 59.9 | 35.8 | 57.2 | 67.4 | – | Paper | Code | 2021 | linked, not harvested | report | ||||
| 58 | DyHead (ResNeXt-64x4d-101-DCN, multi scale) | 54 | 72.1 | 59.3 | – | Paper | Code | 2021 | 7 of 9 ran · 2 unverified | report | |||||||
| 59 | dBOT ViT-B (CLIP) | 53.6 | – | Paper | Code | 2022 | 5 of 13 ran · 8 unverified | report | |||||||||
| 60 | PAA (ResNext-152-32x8d + DCN, multi-scale) | 53.5 | 71.6 | 59.1 | 36.0 | 56.3 | 66.9 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 61 | LSNet (Res2Net-101+ DCN, multi-scale) | 53.5 | 71.1 | 59.2 | 35.2 | 56.4 | 65.8 | – | Paper | Code | 2021 | linked, not harvested | report | ||||
| 62 | dBOT ViT-B | 53.5 | – | Paper | Code | 2022 | 5 of 13 ran · 8 unverified | report | |||||||||
| 63 | ResNeSt-200 (multi-scale) | 53.3 | 72.0 | 58.0 | 35.1 | 56.2 | 66.8 | – | Paper | Code | 2020 | 8 of 48 ran · 40 unverified | report | ||||
| 64 | Cascade Mask R-CNN (Triple-ResNeXt152, multi-scale) | 53.3 | 71.9 | 58.5 | 35.5 | 55.8 | 66.7 | – | Paper | Code | 2019 | 0 of 4 ran · 4 unverified | report | ||||
| 65 | DetectoRS (ResNeXt-101-32x4d, single-scale) | 53.3 | 71.6 | 58.5 | 33.9 | 56.5 | 66.9 | – | Paper | Code | 2020 | 0 of 4 ran · 4 unverified | report | ||||
| 66 | GFLV2 (Res2Net-101, DCN, multiscale) | 53.3 | 70.9 | 59.2 | 35.7 | 56.1 | 65.6 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 67 | YOLOv7-X (114 fps) | 53.1 | ✓ | Paper | Code | 2022 | 1 of 11 ran · 10 unverified | report | |||||||||
| 68 | RelationNet++ (ResNeXt-64x4d-101-DCN) | 52.7 | – | Paper | Code | 2020 | 0 of 5 ran · 5 unverified | report | |||||||||
| 69 | EfficientDet-D7 (1536) | 52.6 | 71.6 | 56.9 | ✓ | Paper | Code | 2019 | 11 of 70 ran · 59 unverified | report | |||||||
| 70 | YOLOv4-P5 with TTA | 52.5 | 70.3 | 58 | – | Paper | Code | 2020 | 1 of 1 ran · 0 unverified | report | |||||||
| 71 | Deformable DETR (ResNeXt-101+DCN) | 52.3 | 71.9 | 58.1 | 34.4 | 54.4 | 65.6 | 17G | 17.3G | – | Paper | Code | 2020 | 29 of 55 ran · 26 unverified | report | ||
| 72 | GCNet (ResNeXt-101 + DCN + cascade + GC r4) | 52.3 | 70.9 | 56.9 | – | Paper | Code | 2020 | linked, not harvested | report | |||||||
| 73 | PP-YOLOE-x(CSPRepResNet-x, 640x640, single-scale ) | 52.2 | 69.9 | 56.5 | 33.3 | 56.3 | 66.4 | – | Paper | Code | 2022 | 5 of 27 ran · 22 unverified | report | ||||
| 74 | RetinaNet (SpineNet-190, 1280x1280) | 52.1 | 71.8 | 56.5 | 35.4 | 55 | 63.6 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 75 | RepPoints v2 (ResNeXt-101, DCN, multi-scale) | 52.1 | 70.1 | 57.5 | 34.5 | 54.6 | 63.6 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 76 | AC-FPN Cascade R-CNN (X-152-32x8d-FPN-IN5k, multi scale, only CEM) | 51.9 | 70.4 | 57 | 34.2 | 54.8 | 64.7 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 77 | OTA (ResNeXt-101+DCN, multiscale) | 51.5 | 68.6 | 57.1 | 34.1 | 53.7 | 64.1 | – | Paper | Code | 2021 | 3 of 3 ran · 0 unverified | report | ||||
| 78 | YOLOX-x(Modified CSP v5, 640x640, single-scale) | 51.5 | ✓ | Paper | Code | 2021 | 1 of 23 ran · 22 unverified | report | |||||||||
| 79 | PP-YOLOE-l(CSPRepResNet-l, 640x640, single-scale ) | 51.4 | 68.9 | 55.6 | 31.4 | 55.3 | 66.1 | – | Paper | Code | 2022 | 5 of 27 ran · 22 unverified | report | ||||
| 80 | YOLOv7 (161 fps) | 51.4 | ✓ | Paper | Code | 2022 | 1 of 11 ran · 10 unverified | report | |||||||||
| 81 | UniverseNet-20.08d (Res2Net-101, DCN, single-scale) | 51.3 | 70.0 | 55.8 | 31.7 | 55.3 | 64.9 | – | Paper | Code | 2021 | linked, not harvested | report | ||||
| 82 | TSD(SENet154-DCN,multi-scale) | 51.2 | 71.9 | 56.0 | 33.8 | 54.8 | 64.2 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 83 | YOLOX-X (Modified CSP v5) | 51.2 | 69.6 | 55.7 | 31.2 | 56.1 | 66.1 | 99.1 | – | Paper | Code | 2021 | 1 of 23 ran · 22 unverified | report | |||
| 84 | iBOT (ViT-B/16) | 51.2 | – | Paper | Code | 2021 | 0 of 1 ran · 1 unverified | report | |||||||||
| 85 | RetinaNet (SpineNet-143, 1280x1280) | 50.7 | 70.4 | 54.9 | 33.6 | 53.9 | 62.1 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 86 | ATSS (ResNetXt-64x4d-101+DCN,multi-scale) | 50.7 | 68.9 | 56.3 | 33.2 | 52.9 | 62.4 | – | Paper | Code | 2019 | 0 of 4 ran · 4 unverified | report | ||||
| 87 | NAS-FPN (AmoebaNet-D, learned aug) | 50.7 | 34.2 | 55.5 | 64.5 | – | Paper | Code | 2019 | 3 of 3 ran · 0 unverified | report | ||||||
| 88 | Boosting R-CNN* | 50.7 | – | Paper | Code | 2022 | linked, not harvested | report | |||||||||
| 89 | GFLV2 (Res2Net-101, DCN) | 50.6 | 69 | 55.3 | 31.3 | 54.3 | 63.5 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 90 | aLRP Loss (ResNext-101-64x4d, DCN, multiscale test) | 50.2 | 70.3 | 53.9 | 32.0 | 53.1 | 63.0 | – | Paper | Code | 2020 | 2 of 2 ran · 0 unverified | report | ||||
| 91 | FreeAnchor + SEPC (DCN, ResNext-101-64x4d) | 50.1 | 69.8 | 54.3 | 31.3 | 53.3 | 63.7 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 92 | D2Det (ResNet-101-DCN, multi-scale test) | 50.1 | 69.4 | 54.9 | 32.7 | 52.7 | 62.1 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 93 | Dynamic R-CNN (ResNet-101-DCN, multi-scale) | 50.1 | 68.3 | 55.6 | 32.8 | 53.0 | 61.2 | – | Paper | Code | 2020 | 2 of 18 ran · 16 unverified | report | ||||
| 94 | TSD(ResNet-101-Deformable, Image Pyramid) | 49.4 | 69.6 | 54.4 | 32.7 | 52.5 | 61.0 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 95 | RepPoints v2 (ResNeXt-101, DCN) | 49.4 | 68.9 | 53.4 | 30.3 | 52.1 | 62.3 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 96 | A2MIM (ViT-B) | 49.4 | – | Paper | Code | 2022 | linked, not harvested | report | |||||||||
| 97 | iBOT (ViT-S/16) | 49.4 | – | Paper | Code | 2021 | 0 of 1 ran · 1 unverified | report | |||||||||
| 98 | CPNDet (Hourglass-104, multi-scale) | 49.2 | 67.3 | 53.7 | 31.0 | 51.9 | 62.4 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 99 | GFLV2 (ResNeXt-101, 32x4d, DCN) | 49 | 67.6 | 53.5 | 29.7 | 52.4 | 61.4 | 3G | – | Paper | Code | 2020 | linked, not harvested | report | |||
| 100 | aLRP Loss (ResNext-101-64x4d, DCN, single scale) | 48.9 | 69.3 | 52.5 | 30.8 | 51.5 | 62.1 | – | Paper | Code | 2020 | 2 of 2 ran · 0 unverified | report | ||||
| 101 | PP-YOLOE-m(CSPRepResNet-m, 640x640, single-scale ) | 48.9 | 66.5 | 53.0 | 28.6 | 52.9 | 63.8 | – | Paper | Code | 2022 | 5 of 27 ran · 22 unverified | report | ||||
| 102 | UniverseNet-20.08 (Res2Net-50, DCN, single-scale) | 48.8 | 67.5 | 53.0 | 30.1 | 52.3 | 61.1 | – | Paper | Code | 2021 | linked, not harvested | report | ||||
| 103 | SOLQ (ResNet101, single scale) | 48.7 | – | Paper | Code | 2021 | linked, not harvested | report | |||||||||
| 104 | RetinaNet (SpineNet-96, 1024x1024) | 48.6 | 68.4 | 52.5 | 32 | 52.3 | 62 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 105 | TridentNet (ResNet-101-Deformable, Image Pyramid) | 48.4 | 69.7 | 53.5 | 31.8 | 51.3 | 60.3 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 106 | GCNet (ResNeXt-101 + DCN + cascade + GC r4) | 48.4 | 67.6 | 52.7 | 54.8G | – | Paper | Code | 2019 | linked, not harvested | report | ||||||
| 107 | GFLV2 (ResNet-101-DCN) | 48.3 | 66.5 | 52.8 | 28.8 | 51.9 | 60.7 | 3G | – | Paper | Code | 2020 | linked, not harvested | report | |||
| 108 | Swin-S (RPE w/ GAB) | 48.23 | – | Paper | Code | 2023 | linked, not harvested | report | |||||||||
| 109 | GFL (X-101-32x4d-DCN, single-scale) | 48.2 | 67.4 | 52.6 | 29.2 | 51.7 | 60.2 | – | Paper | Code | 2020 | 1 of 24 ran · 23 unverified | report | ||||
| 110 | ISTR (ResNet101-FPN-3x, single-scale) | 48.1 | 28.7 | 50.4 | 61.5 | – | Paper | Code | 2021 | linked, not harvested | report | ||||||
| 111 | YOLOX-Darknet53(Darknet53, 640x640, single-scale) | 48.0 | ✓ | Paper | Code | 2021 | 1 of 23 ran · 22 unverified | report | |||||||||
| 112 | DAT-S (RetinaNet) | 47.9 | 69.6 | 51.2 | 32.3 | 51.8 | 63.4 | – | Paper | Code | 2022 | 8 of 10 ran · 2 unverified | report | ||||
| 113 | aLRP Loss (ResNext-101-64x4d, single scale) | 47.8 | 68.4 | 51.1 | 30.2 | 50.8 | 59.1 | – | Paper | Code | 2020 | 2 of 2 ran · 0 unverified | report | ||||
| 114 | MatrixNet Corners (ResNet-152, multi-scale) | 47.8 | 66.2 | 52.3 | 29.7 | 50.4 | 60.7 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 115 | SOLQ (ResNet50, single scale) | 47.8 | – | Paper | Code | 2021 | linked, not harvested | report | |||||||||
| 116 | DyHead (ResNeXt-64x4d-101) | 47.7 | 65.7 | 51.9 | – | Paper | Code | 2021 | 7 of 9 ran · 2 unverified | report | |||||||
| 117 | SAPD (ResNeXt-101, single-scale) | 47.4 | 67.4 | 51.1 | 28.1 | 50.3 | 61.5 | – | Paper | Code | 2019 | 0 of 6 ran · 6 unverified | report | ||||
| 118 | PANet (ResNeXt-101, multi-scale) | 47.4 | 67.2 | 51.8 | 30.1 | 51.7 | 60.0 | – | Paper | Code | 2018 | 1 of 4 ran · 3 unverified | report | ||||
| 119 | HTC (HRNetV2p-W48) | 47.3 | 65.9 | 51.2 | 28.0 | 49.7 | 59.8 | 15G | 71.7G | – | Paper | Code | 2019 | 3 of 34 ran · 31 unverified | report | ||
| 120 | HTC (ResNeXt-101-FPN) | 47.1 | 63.9 | 44.7 | 22.8 | 43.9 | 54.6 | – | Paper | Code | 2019 | 0 of 11 ran · 11 unverified | report | ||||
| 121 | CenterNet511 (Hourglass-104, multi-scale) | 47.0 | 64.5 | 50.7 | 28.9 | 49.9 | 58.9 | – | Paper | Code | 2019 | 2 of 11 ran · 9 unverified | report | ||||
| 122 | MAL (ResNeXt101, multi-scale) | 47.0 | – | Paper | Code | 2019 | 1 of 10 ran · 9 unverified | report | |||||||||
| 123 | ISTR (ResNet50-FPN-3x) | 46.8 | – | Paper | Code | 2021 | linked, not harvested | report | |||||||||
| 124 | RetinaNet (SpineNet-49, 896x896) | 46.7 | 66.3 | 50.6 | 29.1 | 50.1 | 61.7 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 125 | RPDet (ResNet-101-DCN, multi-scale) | 46.5 | 67.4 | 50.9 | 30.3 | 49.7 | 57.1 | – | Paper | Code | 2019 | 1 of 3 ran · 2 unverified | report | ||||
| 126 | HoughNet (MS) | 46.4 | 65.1 | 50.7 | 29.1 | 48.5 | 58.1 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 127 | PPDet (ResNeXt-101-FPN, multiscale) | 46.3 | 64.8 | 51.6 | 31.4 | 49.9 | 56.4 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 128 | GFLV2 (ResNet-101) | 46.2 | 64.3 | 50.5 | 27.8 | 49.9 | 57 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 129 | SNIPER (ResNet-101) | 46.1 | 67.0 | 51.6 | 29.6 | 48.9 | 58.1 | 29G | – | Paper | Code | 2018 | linked, not harvested | report | |||
| 130 | Mask R-CNN (HRNetV2p-W48 + cascade) | 46.1 | 64.0 | 50.3 | 27.1 | 48.6 | 58.3 | 15G | 61.8G | – | Paper | Code | 2019 | 3 of 34 ran · 31 unverified | report | ||
| 131 | ResNeXt-64x4d-101 NAS-FCOS @128-256 w/improvements | 46.1 | – | Paper | Code | 2019 | 0 of 7 ran · 7 unverified | report | |||||||||
| 132 | DCNv2 (ResNet-101, multi-scale) | 46.0 | 67.9 | 50.8 | 27.8 | 49.1 | 59.5 | – | Paper | Code | 2018 | 2 of 13 ran · 11 unverified | report | ||||
| 133 | Gaussian-FCOS | 46 | – | Paper | – | 2020 | no code linked | report | |||||||||
| 134 | Cascade R-CNN-FPN (ResNet-101, map-guided) | 45.9 | 64.2 | 50 | 26.3 | 49 | 58.6 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 135 | MAL (ResNeXt101, single-scale) | 45.9 | – | Paper | Code | 2019 | 1 of 10 ran · 9 unverified | report | |||||||||
| 136 | CenterMask+VoVNetV2-99 (single-scale) | 45.8 | 64.5 | 27.8 | 48.3 | 57.6 | – | Paper | Code | 2019 | 1 of 2 ran · 1 unverified | report | |||||
| 137 | D-RFCN + SNIP (DPN-98 with flip, multi-scale) | 45.7 | 67.3 | 51.1 | 29.3 | 48.8 | 57.1 | – | Paper | – | 2017 | no code linked | report | ||||
| 138 | YOLOv4 (CD53) | 45.5 | 64.1 | 49.5 | 27 | 49 | 56.7 | ✓ | Paper | Code | 2020 | 1 of 1 ran · 0 unverified | report | ||||
| 139 | AC-FPN Cascade R-CNN(ResNet-101, single scale) | 45 | 64.4 | 49 | 26.9 | 47.7 | 56.6 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 140 | FreeAnchor (ResNeXt-101) | 44.8 | 64.3 | 48.4 | 27 | 47.9 | 56 | – | Paper | Code | 2019 | 3 of 3 ran · 0 unverified | report | ||||
| 141 | FCOS (ResNeXt-64x4d-101-FPN 4 + improvements) | 44.7 | 64.1 | 48.4 | 27.6 | 47.5 | 55.6 | – | Paper | Code | 2019 | 13 of 40 ran · 27 unverified | report | ||||
| 142 | CenterMask+VoVNet2-57 (single-scale) | 44.7 | 63.1 | 48.6 | 27.1 | 55.9 | – | Paper | Code | 2019 | 1 of 2 ran · 1 unverified | report | |||||
| 143 | FSAF (ResNeXt-101, multi-scale) | 44.6 | 65.2 | 48.6 | 29.7 | 47.1 | 54.6 | – | Paper | Code | 2019 | 0 of 8 ran · 8 unverified | report | ||||
| 144 | aLRP Loss (ResNext-101, DCN, 500 scale) | 44.6 | 65.0 | 47.5 | 24.6 | 48.1 | 58.3 | – | Paper | Code | 2020 | 2 of 2 ran · 0 unverified | report | ||||
| 145 | CenterMask + X-101-32x8d (single-scale) | 44.6 | 63.4 | 48.4 | 47.2 | – | Paper | Code | 2019 | 1 of 2 ran · 1 unverified | report | ||||||
| 146 | RetinaNet (SpineNet-49, 640x640) | 44.3 | 63.8 | 47.6 | 25.9 | 47.7 | 61.1 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 147 | YOLOF-DC5 | 44.3 | 62.9 | 47.5 | 24.0 | 48.5 | 60.4 | – | Paper | Code | 2021 | linked, not harvested | report | ||||
| 148 | GFLV2 (ResNet-50) | 44.3 | 62.3 | 48.5 | 26.8 | 47.7 | 54.1 | – | Paper | Code | 2020 | linked, not harvested | report | ||||
| 149 | InterNet (ResNet-101-FPN, multi-scale) | 44.2 | 67.5 | 51.1 | 27.2 | 50.3 | 57.7 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 150 | M2Det (VGG-16, multi-scale) | 44.2 | 64.6 | 49.3 | 29.2 | 47.9 | 55.1 | 34G | – | Paper | Code | 2018 | 1 of 1 ran · 0 unverified | report | |||
| 151 | Faster R-CNN (LIP-ResNet-101-MD w FPN) | 43.9 | 65.7 | 48.1 | 25.4 | 46.7 | 56.3 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 152 | M2Det (ResNet-101, multi-scale) | 43.9 | 64.4 | 48 | 29.6 | 49.6 | 54.3 | 27G | – | Paper | Code | 2018 | 1 of 1 ran · 0 unverified | report | |||
| 153 | YOLOv3 @800 + ASFF* (Darknet-53) | 43.9 | 64.1 | 49.2 | 27.0 | 46.6 | 53.4 | ✓ | Paper | Code | 2019 | linked, not harvested | report | ||||
| 154 | FoveaBox (ResNeXt-101) | 43.9 | 63.5 | 47.7 | 26.8 | 46.9 | 55.6 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 155 | ExtremeNet (Hourglass-104, multi-scale) | 43.7 | 60.5 | 47.0 | 24.1 | 46.9 | 57.6 | 180G | – | Paper | Code | 2019 | 1 of 5 ran · 4 unverified | report | |||
| 156 | YOLOv4-608 | 43.5 | 65.7 | 47.3 | 26.7 | 46.7 | 53.3 | ✓ | Paper | Code | 2020 | 24 of 184 ran · 160 unverified | report | ||||
| 157 | SNIPER (ResNet-50) | 43.5 | 65.0 | 48.6 | 26.1 | 46.3 | 56.0 | 29G | – | Paper | Code | 2018 | linked, not harvested | report | |||
| 158 | CenterNet (HRNetV2-W48) | 43.5 | 46.5 | 22.2 | 57.8 | 16G | 21.7G | – | Paper | Code | 2019 | 3 of 34 ran · 31 unverified | report | ||||
| 159 | D-RFCN + SNIP (ResNet-101, multi-scale) | 43.4 | 65.5 | 48.4 | 27.2 | 46.5 | 54.9 | – | Paper | – | 2017 | no code linked | report | ||||
| 160 | Grid R-CNN (ResNeXt-101-FPN) | 43.2 | 63.0 | 46.6 | 25.1 | 46.5 | 55.2 | – | Paper | Code | 2018 | 0 of 3 ran · 3 unverified | report | ||||
| 161 | FCOS (ResNeXt-101-64x4d-FPN) | 43.2 | 62.8 | 46.6 | 26.5 | 46.2 | 53.3 | – | Paper | Code | 2019 | 13 of 40 ran · 27 unverified | report | ||||
| 162 | CornerNet-Saccade (Hourglass-104, multi-scale) | 43.2 | 24.4 | 44.6 | 57.3 | – | Paper | Code | 2019 | 1 of 27 ran · 26 unverified | report | ||||||
| 163 | PP-YOLOE-s(CSPRepResNet-s, 640x640, single-scale ) | 43.1 | 60.5 | 46.6 | 23.2 | 46.4 | 56.9 | – | Paper | Code | 2022 | 5 of 27 ran · 22 unverified | report | ||||
| 164 | Libra R-CNN (ResNeXt-101-FPN) | 43.0 | 64 | 47 | 25.3 | 45.6 | 54.6 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 165 | DyHead (ResNet-50) | 43 | 60.7 | 46.8 | – | Paper | Code | 2021 | 7 of 9 ran · 2 unverified | report | |||||||
| 166 | RPDet (ResNet-101-DCN) | 42.8 | 65.0 | 46.3 | 24.9 | 46.2 | 54.7 | – | Paper | Code | 2019 | 1 of 3 ran · 2 unverified | report | ||||
| 167 | SpineNet-49 (640, RetinaNet, single-scale) | 42.8 | 62.3 | 46.1 | 23.7 | 45.2 | 57.3 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 168 | Cascade R-CNN (ResNet-101-FPN+, cascade) | 42.8 | 62.1 | 46.3 | 23.7 | 45.5 | 55.2 | – | Paper | Code | 2017 | 2 of 2 ran · 0 unverified | report | ||||
| 169 | Cascade R-CNN | 42.8 | 62.1 | 46.3 | 23.7 | 45.5 | 55.2 | 15G | – | Paper | Code | 2019 | linked, not harvested | report | |||
| 170 | TridentNet (ResNet-101) | 42.7 | 63.6 | 46.5 | 23.9 | 46.6 | 56.6 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 171 | FCOS (ResNeXt-32x8d-101-FPN) | 42.7 | 62.2 | 46.1 | 26.0 | 45.6 | 52.6 | – | Paper | Code | 2019 | 13 of 40 ran · 27 unverified | report | ||||
| 172 | RetinaMask (ResNeXt-101-FPN-GN) | 42.6 | 62.5 | 46.0 | 24.8 | 45.6 | 53.8 | 12G | – | Paper | Code | 2019 | linked, not harvested | report | |||
| 173 | TAL + TAP | 42.5 | 60.3 | 46.4 | – | Paper | Code | 2021 | 2 of 4 ran · 2 unverified | report | |||||||
| 174 | Faster R-CNN (HRNetV2p-W48) | 42.4 | 63.6 | 46.4 | 24.9 | 44.6 | 53.0 | 16G | 20.8G | – | Paper | Code | 2019 | 3 of 34 ran · 31 unverified | report | ||
| 175 | HSD (Rest101, 768x768, single-scale test) | 42.3 | 61.2 | 46.9 | 22.8 | 47.3 | 55.9 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 176 | CornerNet511 (Hourglass-104, multi-scale) | 42.1 | 57.8 | 45.3 | 20.8 | 44.8 | 56.7 | – | Paper | Code | 2018 | 3 of 11 ran · 8 unverified | report | ||||
| 177 | FoveaBox (ResNeXt-101) | 42.1 | – | Paper | Code | 2019 | linked, not harvested | report | |||||||||
| 178 | FCOS (HRNet-W32-5l) | 42.0 | 60.4 | 45.3 | 25.4 | 45.0 | 51.0 | – | Paper | Code | 2019 | 13 of 40 ran · 27 unverified | report | ||||
| 179 | FoveaBox (ResNeXt-101) | 41.9 | – | Paper | Code | 2019 | linked, not harvested | report | |||||||||
| 180 | RefineDet512+ (ResNet-101) | 41.8 | 62.9 | 45.7 | 25.6 | 45.1 | 54.1 | – | Paper | Code | 2017 | linked, not harvested | report | ||||
| 181 | GHM-C + GHM-R (RetinaNet-FPN-ResNeXt-101) | 41.6 | 62.8 | 44.2 | 22.3 | 45.1 | 55.3 | – | Paper | Code | 2018 | 0 of 1 ran · 1 unverified | report | ||||
| 182 | CenterNet-DLA (DLA-34, multi-scale) | 41.6 | 21.5 | 43.9 | 56.0 | 26G | – | Paper | Code | 2019 | 10 of 130 ran · 120 unverified | report | |||||
| 183 | RetinaNet (SpineNet-49S, 640x640) | 41.5 | 60.5 | 44.6 | 23.3 | 45 | 58 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 184 | RPDet (ResNet-101) | 41 | 62.9 | 44.3 | 23.6 | 44.1 | 51.7 | – | Paper | Code | 2019 | 1 of 3 ran · 2 unverified | report | ||||
| 185 | M2Det (VGG-16, single-scale) | 41.0 | 59.7 | 45 | 22.1 | 46.5 | 53.8 | 34G | – | Paper | Code | 2018 | 1 of 1 ran · 0 unverified | report | |||
| 186 | LeYOLO (Large@768) | 41.0 | 2.4 | 8.4 | – | Paper | Code | 2024 | linked, not harvested | report | |||||||
| 187 | FSAF (ResNet-101, single-scale) | 40.9 | 61.5 | 44 | 24 | 44.2 | 51.3 | 38G | – | Paper | Code | 2019 | 0 of 8 ran · 8 unverified | report | |||
| 188 | RetinaNet (ResNeXt-101-FPN) | 40.8 | 61.1 | 44.1 | 24.1 | 44.2 | 51.2 | 4G | – | Paper | Code | 2017 | 11 of 11 ran · 0 unverified | report | |||
| 189 | Cascade R-CNN (ResNet-50-FPN+, cascade) | 40.6 | 59.9 | 44 | 22.6 | 42.7 | 52.1 | 12G | – | Paper | Code | 2017 | 2 of 2 ran · 0 unverified | report | |||
| 190 | Faster R-CNN (Cascade RPN) | 40.6 | 58.9 | 44.5 | 22.0 | 42.8 | 52.6 | 5G | ✓ | Paper | Code | 2019 | linked, not harvested | report | |||
| 191 | ResNet-50-DW-DPN (Deformable Kernels) | 40.6 | 24.6 | 43.9 | 53.3 | – | Paper | Code | 2019 | linked, not harvested | report | ||||||
| 192 | IoU-Net | 40.6 | – | Paper | Code | 2018 | 2 of 18 ran · 16 unverified | report | |||||||||
| 193 | FCOS (HRNetV2p-W48) | 40.5 | 59.3 | 23.4 | 42.6 | 51.0 | 16G | 27.3G | ✓ | Paper | Code | 2019 | 3 of 34 ran · 31 unverified | report | |||
| 194 | ResNet-50-FPN Mask R-CNN + KL Loss + var voting + soft-NMS | 40.4 | – | Paper | Code | 2018 | 0 of 4 ran · 4 unverified | report | |||||||||
| 195 | RDSNet (ResNet-101, RetinaNet, mask, MBRM) | 40.3 | 60.1 | 43 | 22.1 | 43.5 | 51.5 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 196 | ExtremeNet (Hourglass-104, single-scale) | 40.2 | 55.5 | 43.2 | 20.4 | 43.2 | 53.1 | 180G | – | Paper | Code | 2019 | 1 of 5 ran · 4 unverified | report | |||
| 197 | Mask R-CNN (ResNet-101-FPN, CBN) | 40.1 | 60.5 | 44.1 | 35.8 | 57.3 | 38.5 | – | Paper | Code | 2020 | 0 of 6 ran · 6 unverified | report | ||||
| 198 | Fast R-CNN (Cascade RPN) | 40.1 | 59.4 | 43.8 | 22.1 | 42.4 | 51.6 | 5G | ✓ | Paper | Code | 2019 | linked, not harvested | report | |||
| 199 | Mask R-CNN (ResNeXt-101-FPN) | 39.8 | 62.3 | 43.4 | 22.1 | 43.2 | 51.2 | 9G | – | Paper | Code | 2017 | 42 of 140 ran · 98 unverified | report | |||
| 200 | GA-Faster-RCNN | 39.8 | 59.2 | 43.5 | 21.8 | 42.6 | 50.7 | – | Paper | Code | 2019 | linked, not harvested | report | ||||
| 201 | ResNet-50 NAS-FCOS @256 | 39.8 | – | Paper | Code | 2019 | 0 of 7 ran · 7 unverified | report | |||||||||
| 202 | A2MIM (ResNet-50 2x) | 39.8 | – | Paper | Code | 2022 | linked, not harvested | report | |||||||||
| 203 | FPN (ResNet101 backbone) | 39.5 | – | Paper | Code | 2017 | linked, not harvested | report | |||||||||
| 204 | RetinaMask (ResNet-50-FPN) | 39.4 | 58.6 | 42.3 | 21.9 | 42.0 | 51.0 | 9G | – | Paper | Code | 2019 | linked, not harvested | report | |||
| 205 | LeYOLO (Medium@640) | 39.3 | 5.8 | – | Paper | Code | 2024 | linked, not harvested | report | ||||||||
| 206 | AA-ResNet-10 + RetinaNet | 39.2 | 24.5G | – | Paper | Code | 2019 | 3 of 6 ran · 3 unverified | report | ||||||||
| 207 | MAL (ResNet50, single-scale) | 39.2 | – | Paper | Code | 2019 | 1 of 10 ran · 9 unverified | report | |||||||||
| 208 | RetinaNet (ResNet-101-FPN) | 39.1 | 59.1 | 42.3 | 21.8 | 42.7 | 50.2 | 4G | – | Paper | Code | 2017 | 11 of 11 ran · 0 unverified | report | |||
| 209 | Cascade R-CNN (ResNet-101-FPN+) | 38.8 | 61.1 | 41.9 | 21.3 | 41.8 | 49.8 | 3G | – | Paper | Code | 2017 | 2 of 2 ran · 0 unverified | report | |||
| 210 | M2Det (ResNet-101, single-scale) | 38.8 | 59.4 | 41.7 | 20.5 | 43.9 | 53.4 | 27G | – | Paper | Code | 2018 | 1 of 1 ran · 0 unverified | report | |||
| 211 | SaccadeNet (DLA-34-DCN) | 38.5 | 55.6 | 41.4 | 19.2 | 42.1 | 50.6 | 46G | – | Paper | Code | 2020 | linked, not harvested | report | |||
| 212 | Mask R-CNN (ResNet-101-FPN) | 38.2 | 60.3 | 41.7 | 20.1 | 41.1 | 50.2 | 9G | – | Paper | Code | 2017 | 42 of 140 ran · 98 unverified | report | |||
| 213 | LeYOLO (Small@640) | 38.2 | 1.9 | 4.51 | – | Paper | Code | 2024 | linked, not harvested | report | |||||||
| 214 | WSMA-Seg | 38.1 | – | Paper | – | 2019 | no code linked | report | |||||||||
| 215 | Faster R-CNN + FPN + CGD | 37.9 | – | Paper | Code | 2019 | linked, not harvested | report | |||||||||
| 216 | CornerNet511 (Hourglass-52, single-scale) | 37.8 | 53.7 | 40.1 | 17.0 | 39.0 | 50.5 | – | Paper | Code | 2018 | 3 of 11 ran · 8 unverified | report | ||||
| 217 | RefineDet512+ (VGG-16) | 37.6 | 58.7 | 40.8 | 22.7 | 40.3 | 48.3 | – | Paper | Code | 2017 | linked, not harvested | report | ||||
| 218 | DeformConv-R-FCN (Aligned-Inception-ResNet) | 37.5 | 58.0 | 19.4 | 40.1 | 52.5 | – | Paper | Code | 2017 | 3 of 9 ran · 6 unverified | report | |||||
| 219 | Faster R-CNN (ImageNet+300M) | 37.4 | 58 | 40.1 | 17.5 | 41.1 | 51.2 | – | Paper | Code | 2017 | linked, not harvested | report | ||||
| 220 | Mask R-CNN (Bottleneck-injected ResNet-50, FPN) | 36.9 | – | Paper | Code | 2020 | 10 of 26 ran · 16 unverified | report | |||||||||
| 221 | Faster R-CNN + TDM | 36.8 | – | Paper | Code | 2016 | linked, not harvested | report | |||||||||
| 222 | Cascade R-CNN (ResNet-50-FPN+) | 36.5 | 59 | 39.2 | 20.3 | 38.8 | 46.4 | 3G | – | Paper | Code | 2017 | 2 of 2 ran · 0 unverified | report | |||
| 223 | RefineDet512 (ResNet-101) | 36.4 | 57.5 | 39.5 | 16.6 | 39.9 | 51.4 | – | Paper | Code | 2017 | linked, not harvested | report | ||||
| 224 | Faster R-CNN + FPN | 36.2 | 2G | ✓ | Paper | Code | 2016 | 16 of 51 ran · 35 unverified | report | ||||||||
| 225 | Faster R-CNN (Bottleneck-injected ResNet-50 and FPN) | 35.9 | – | Paper | Code | 2020 | 10 of 26 ran · 16 unverified | report |
All 225 rows shown. 225 link to a paper page on this site; 13 are marked as using additional training data in the archive. No GitHub stars are tracked; "Code" is the first repository the archive lists for the row. The archive carries no row tags, review links or community-submitted rows for this table; none are shown. archive 2025-07-28
Syntology Ran reads "N of M ran · U unverified": of the M code samples Syntology harvested from repositories linked to that row's paper (joined by arXiv id), N executed on a synthesized input and the other U = M−N are unverified (harvested, no recorded run). It counts code from repositories linked to that row's paper, not this result: the row's number was not reproduced and nothing here is a correctness claim. The other cell texts mean no graph line for the row: "linked, not harvested" (the archive links code, Syntology has not harvested it), "no code linked" (no code link in the archive), "not matched" (the row's paper URL matched no paper on this site). 131 rows have a graph line, from 72 distinct papers; 110 rows (57 papers) have at least one sample that ran. Counting each paper once: Syntology ran 405 of 1,454 samples; 1,049 unverified. Separately, 224 of those 1,454 are pointer-only (licence): the site points at that code rather than redistributing it, a licence property recorded for ran and unverified samples alike; each cell's tooltip carries the row's own pointer-only count. Read from the graph 2026-09-24. Per-sample status is on the paper page.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections