| TinyFormer: Preserving Tiny Objects in YOLO-DETR Hybrid Real-time Detectors | ✓ Link | 62.5 | 126 (RTX 3090) | TinyFormer-XL-PBM (Objects365 pre-training) | 2026-05-24 |
| TinyFormer: Preserving Tiny Objects in YOLO-DETR Hybrid Real-time Detectors | ✓ Link | 60.6 | 126 (RTX 3090) | TinyFormer-XL-PBM | 2026-05-24 |
| RF-DETR: Neural Architecture Search for Real-Time Detection Transformers | | 60.1 | 58 (T4) | RF-DETR-2XL (880px) (Objects365 pre-training) | 2025-11-12 |
| DEIM: DETR with Improved Matching for Fast Convergence | ✓ Link | 59.5 | 78 (T4) | DEIM-D-FINE-X+ (Objects365 pre-training) | 2024-12-05 |
| D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement | ✓ Link | 59.3 | 78 (T4) | D-FINE-X+ (Objects365 pre-training) | 2024-10-17 |
| RF-DETR: Neural Architecture Search for Real-Time Detection Transformers | | 58.6 | 87 (T4) | RF-DETR-XL (700px) (Objects365 pre-training) | 2025-11-12 |
| TinyFormer: Preserving Tiny Objects in YOLO-DETR Hybrid Real-time Detectors | ✓ Link | 58.4 | 216 (RTX 3090) | TinyFormer-X | 2026-05-24 |
| LW-DETR: A Transformer Replacement to YOLO for Real-Time Detection | ✓ Link | 58.3 | 52 (T4) | LW-DETR-xlarge (Objects365 pre-training) | 2024-06-05 |
| Real-Time Object Detection Meets DINOv3 | ✓ Link | 57.8 | | DEIMv2-X (DINOv3 backbone) | 2025-09-25 |
| Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models | ✓ Link | 57.5 | 85 (T4) | YOLO26x (Objects365 pre-training) | 2026-06-02 |
| YOLOv6 v3.0: A Full-Scale Reloading | ✓ Link | 57.2 | 26 (T4) | YOLOv6-L6 (1280, self-distillation) | 2023-01-13 |
| D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement | ✓ Link | 57.1 | 124 (T4) | D-FINE-L+ (Objects365 pre-training) | 2024-10-17 |
| RT-DETRv4: Painlessly Furthering Real-Time Object Detection with Vision Foundation Models | | 57.0 | 78 (T4) | RT-DETRv4-X | 2025-10-29 |
| DEIM: DETR with Improved Matching for Fast Convergence | ✓ Link | 56.5 | 78 (T4) | DEIM-D-FINE-X | 2024-12-05 |
| RF-DETR: Neural Architecture Search for Real-Time Detection Transformers | | 56.5 | 147 (T4) | RF-DETR-L (704px) (Objects365 pre-training) | 2025-11-12 |
| TinyFormer: Preserving Tiny Objects in YOLO-DETR Hybrid Real-time Detectors | ✓ Link | 56.5 | 282 (RTX 3090) | TinyFormer-L | 2026-05-24 |
| VajraV1 -- The most accurate Real Time Object Detector of the YOLO family | ✓ Link | 56.2 | 312 (RTX 4090) | VajraV1-Xlarge | 2025-12-15 |
| LW-DETR: A Transformer Replacement to YOLO for Real-Time Detection | ✓ Link | 56.1 | 114 (T4) | LW-DETR-large (Objects365 pre-training) | 2024-06-05 |
| Real-Time Object Detection Meets DINOv3 | ✓ Link | 56.0 | | DEIMv2-L (DINOv3 backbone) | 2025-09-25 |
| D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement | ✓ Link | 55.8 | 78 (T4) | D-FINE-X | 2024-10-17 |
| YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information | ✓ Link | 55.6 | | YOLOv9-E | 2024-02-21 |
| RT-DETRv4: Painlessly Furthering Real-Time Object Detection with Vision Foundation Models | | 55.4 | 124 (T4) | RT-DETRv4-L | 2025-10-29 |
| YOLOv12: Attention-Centric Real-Time Object Detectors | ✓ Link | 55.2 | 85 (T4) | YOLOv12-X | 2025-02-18 |
| D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement | ✓ Link | 55.1 | 178 (T4) | D-FINE-M+ (Objects365 pre-training) | 2024-10-17 |
| YOLO-PRO: Enhancing Instance-Specific Object Detection with Full-Channel Global Self-Attention | | 55.1 | 75 (T4) | YOLO-PRO-X (YOLO11-based) | 2025-03-04 |
| YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information | ✓ Link | 55.0 | | GELAN-E | 2024-02-21 |
| Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models | ✓ Link | 55.0 | 161 (T4) | YOLO26l (Objects365 pre-training) | 2026-06-02 |
| YOLO-PRO: Enhancing Instance-Specific Object Detection with Full-Channel Global Self-Attention | | 54.9 | 59 (T4) | YOLO-PRO-X (YOLOv8-based) | 2025-03-04 |
| DETRs Beat YOLOs on Real-time Object Detection | ✓ Link | 54.8 | 74 (T4) | RT-DETR-X | 2023-04-17 |
| YOLOv13: Real-Time Object Detection with Hypergraph-Enhanced Adaptive Visual Perception | ✓ Link | 54.8 | 68 (T4) | YOLOv13-X | 2025-06-21 |
| Ultralytics YOLO11 | ✓ Link | 54.7 | 88 (T4) | YOLO11x | 2024-09-27 |
| DEIM: DETR with Improved Matching for Fast Convergence | ✓ Link | 54.7 | 124 (T4) | DEIM-D-FINE-L | 2024-12-05 |
| RF-DETR: Neural Architecture Search for Real-Time Detection Transformers | | 54.7 | 227 (T4) | RF-DETR-M (576px) (Objects365 pre-training) | 2025-11-12 |
| RT-DETRv3: Real-time End-to-End Object Detection with Hierarchical Dense Positive Supervision | ✓ Link | 54.6 | 74 (T4) | RT-DETRv3-R101 | 2024-09-13 |
| YOLOv10: Real-Time End-to-End Object Detection | ✓ Link | 54.4 | 93 (T4) | YOLOv10-X (NMS-free) | 2024-05-23 |
| DETRs Beat YOLOs on Real-time Object Detection | ✓ Link | 54.3 | 74 (T4) | RT-DETR-R101 | 2023-04-17 |
| RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer | ✓ Link | 54.3 | 74 (T4) | RT-DETRv2-X | 2024-07-24 |
| D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement | ✓ Link | 54.0 | 124 (T4) | D-FINE-L | 2024-10-17 |
| Ultralytics YOLOv8 | ✓ Link | 53.9 | 283 (A100) | YOLOv8x | 2023-01-10 |
| YOLO-PRO: Enhancing Instance-Specific Object Detection with Full-Channel Global Self-Attention | | 53.9 | 97 (T4) | YOLO-PRO-L (YOLOv8-based) | 2025-03-04 |
| Hyper-YOLO: When Visual Object Detection Meets Hypergraph Computation | ✓ Link | 53.8 | 73 (T4) | Hyper-YOLO-L | 2024-08-09 |
| DEYO: DETR with YOLO for End-to-End Object Detection | ✓ Link | 53.7 | 65 (T4) | DEYO-X | 2024-02-26 |
| YOLOv12: Attention-Centric Real-Time Object Detectors | ✓ Link | 53.7 | 148 (T4) | YOLOv12-L | 2025-02-18 |
| VajraV1 -- The most accurate Real Time Object Detector of the YOLO family | ✓ Link | 53.7 | 556 (RTX 4090) | VajraV1-Large | 2025-12-15 |
| RT-DETRv4: Painlessly Furthering Real-Time Object Detection with Vision Foundation Models | | 53.5 | 169 (T4) | RT-DETRv4-M | 2025-10-29 |
| HA-DETR: accelerating real-time object detection by replacing decoder self-attention | | 53.5 | 35 (V100) | HA-DETR (ResNet-50) | 2026-04-16 |
| TinyFormer: Preserving Tiny Objects in YOLO-DETR Hybrid Real-time Detectors | ✓ Link | 53.5 | 328 (RTX 3090) | TinyFormer-M | 2026-05-24 |
| RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer | ✓ Link | 53.4 | 108 (T4) | RT-DETRv2-L | 2024-07-24 |
| RT-DETRv3: Real-time End-to-End Object Detection with Hierarchical Dense Positive Supervision | ✓ Link | 53.4 | 109 (T4) | RT-DETRv3-R50 | 2024-09-13 |
| Ultralytics YOLO11 | ✓ Link | 53.4 | 161 (T4) | YOLO11l | 2024-09-27 |
| YOLOv13: Real-Time Object Detection with Hypergraph-Enhanced Adaptive Visual Perception | ✓ Link | 53.4 | 116 (T4) | YOLOv13-L | 2025-06-21 |
| Gold-YOLO: Efficient Object Detector via Gather-and-Distribute Mechanism | ✓ Link | 53.2 | 88 (T4) | Gold-YOLO-L (self-distillation) | 2023-09-20 |
| YOLOv10: Real-Time End-to-End Object Detection | ✓ Link | 53.2 | 137 (T4) | YOLOv10-L (NMS-free) | 2024-05-23 |
| Hyper-YOLO: When Visual Object Detection Meets Hypergraph Computation | ✓ Link | 53.2 | 121 (T4) | Hyper-YOLOv1.1-C | 2024-08-09 |
| Rethinking Features-Fused-Pyramid-Neck for Object Detection | ✓ Link | 53.1 | 97 (V100) | SYOLO (large) | 2025-05-19 |
| Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models | ✓ Link | 53.1 | 213 (T4) | YOLO26m (Objects365 pre-training) | 2026-06-02 |
| DETRs Beat YOLOs on Real-time Object Detection | ✓ Link | 53.0 | 114 (T4) | RT-DETR-L | 2023-04-17 |
| YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information | ✓ Link | 53.0 | | YOLOv9-C | 2024-02-21 |
| Real-Time Object Detection Meets DINOv3 | ✓ Link | 53.0 | | DEIMv2-M (DINOv3-distilled backbone) | 2025-09-25 |
| Ultralytics YOLOv8 | ✓ Link | 52.9 | 418 (A100) | YOLOv8l | 2023-01-10 |
| Octave-YOLO: Cross frequency detection network with octave convolution | | 52.9 | | Octave-YOLO-X | 2024-07-29 |
| RF-DETR: Neural Architecture Search for Real-Time Detection Transformers | | 52.9 | 286 (T4) | RF-DETR-S (512px) (Objects365 pre-training) | 2025-11-12 |
| RTMDet: An Empirical Study of Designing Real-Time Object Detectors | ✓ Link | 52.8 | 323 (RTX 3090) | RTMDet-x | 2022-12-14 |
| YOLOv6 v3.0: A Full-Scale Reloading | ✓ Link | 52.8 | 98 (T4) | YOLOv6-L (self-distillation) | 2023-01-13 |
| DEYO: DETR with YOLO for End-to-End Object Detection | ✓ Link | 52.7 | 100 (T4) | DEYO-L | 2024-02-26 |
| DEIM: DETR with Improved Matching for Fast Convergence | ✓ Link | 52.7 | 180 (T4) | DEIM-D-FINE-M | 2024-12-05 |
| MHAF-YOLO: Multi-Branch Heterogeneous Auxiliary Fusion YOLO for accurate object detection | ✓ Link | 52.7 | | MHAF-YOLO-m | 2025-02-07 |
| VajraV1 -- The most accurate Real Time Object Detector of the YOLO family | ✓ Link | 52.7 | 667 (RTX 4090) | VajraV1-Medium | 2025-12-15 |
| YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information | ✓ Link | 52.5 | | GELAN-C | 2024-02-21 |
| YOLOv10: Real-Time End-to-End Object Detection | ✓ Link | 52.5 | 174 (T4) | YOLOv10-B (NMS-free) | 2024-05-23 |
| LW-DETR: A Transformer Replacement to YOLO for Real-Time Detection | ✓ Link | 52.5 | 179 (T4) | LW-DETR-medium (Objects365 pre-training) | 2024-06-05 |
| YOLOv12: Attention-Centric Real-Time Object Detectors | ✓ Link | 52.5 | 206 (T4) | YOLOv12-M | 2025-02-18 |
| Octave-YOLO: Cross frequency detection network with octave convolution | | 52.3 | | Octave-YOLO-L | 2024-07-29 |
| D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement | ✓ Link | 52.3 | 180 (T4) | D-FINE-M | 2024-10-17 |
| Mamba YOLO: A Simple Baseline for Object Detection with State Space Model | ✓ Link | 52.1 | 233 (RTX 4090) | Mamba YOLO-L | 2024-06-09 |
| P$^2$HCT: Plug-and-Play Hierarchical C2F Transformer for Multi-Scale Feature Fusion | ✓ Link | 52.1 | | YOLOv11-P2HCT-M | 2025-05-19 |
| Hyper-YOLO: When Visual Object Detection Meets Hypergraph Computation | ✓ Link | 52.0 | 111 (T4) | Hyper-YOLO-M | 2024-08-09 |
| RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer | ✓ Link | 51.9 | 145 (T4) | RT-DETRv2-M* | 2024-07-24 |
| RT-DETRv3: Real-time End-to-End Object Detection with Hierarchical Dense Positive Supervision | ✓ Link | 51.7 | 145 (T4) | RT-DETRv3-R50m | 2024-09-13 |
| YOLO-PRO: Enhancing Instance-Specific Object Detection with Full-Channel Global Self-Attention | | 51.6 | 152 (T4) | YOLO-PRO-M (YOLOv8-based) | 2025-03-04 |
| RTMDet: An Empirical Study of Designing Real-Time Object Detectors | ✓ Link | 51.5 | 417 (RTX 3090) | RTMDet-l | 2022-12-14 |
| Ultralytics YOLO11 | ✓ Link | 51.5 | 213 (T4) | YOLO11m | 2024-09-27 |
| YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information | ✓ Link | 51.4 | | YOLOv9-M | 2024-02-21 |
| TinyFormer: Preserving Tiny Objects in YOLO-DETR Hybrid Real-time Detectors | ✓ Link | 51.3 | 429 (RTX 3090) | TinyFormer-S | 2026-05-24 |
| Multi-Branch Auxiliary Fusion YOLO with Re-parameterization Heterogeneous Convolutional for accurate object detection | ✓ Link | 51.2 | | MAF-YOLO-m | 2024-07-05 |
| YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information | ✓ Link | 51.1 | | GELAN-M | 2024-02-21 |
| YOLOv10: Real-Time End-to-End Object Detection | ✓ Link | 51.1 | 211 (T4) | YOLOv10-M (NMS-free) | 2024-05-23 |
| Gold-YOLO: Efficient Object Detector via Gather-and-Distribute Mechanism | ✓ Link | 50.9 | 152 (T4) | Gold-YOLO-M (self-distillation) | 2023-09-20 |
| Real-Time Object Detection Meets DINOv3 | ✓ Link | 50.9 | | DEIMv2-S (DINOv3-distilled backbone) | 2025-09-25 |
| DAMO-YOLO : A Report on Real-Time Object Detection Design | ✓ Link | 50.8 | 126 (V100) | DAMO-YOLO-L | 2022-11-23 |
| DEYO: DETR with YOLO for End-to-End Object Detection | ✓ Link | 50.7 | 140 (T4) | DEYO-M | 2024-02-26 |
| D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement | ✓ Link | 50.7 | 287 (T4) | D-FINE-S+ (Objects365 pre-training) | 2024-10-17 |
| XiYOLO: Energy-Aware Object Detection via Iterative Architecture Search and Scaling | | 50.7 | | XiYOLO-m | 2026-05-07 |
| VajraV1 -- The most accurate Real Time Object Detector of the YOLO family | ✓ Link | 50.4 | 909 (RTX 4090) | VajraV1-Small | 2025-12-15 |
| YOLOv6 v3.0: A Full-Scale Reloading | ✓ Link | 50.3 | 98 (T4) | YOLOv6-S6 (1280) | 2023-01-13 |
| Ultralytics YOLOv8 | ✓ Link | 50.2 | 546 (A100) | YOLOv8m | 2023-01-10 |
| RT-DETRv3: Real-time End-to-End Object Detection with Hierarchical Dense Positive Supervision | ✓ Link | 50.1 | 159 (T4) | RT-DETRv3-R34 | 2024-09-13 |
| YOLOv6 v3.0: A Full-Scale Reloading | ✓ Link | 50.0 | 175 (T4) | YOLOv6-M (self-distillation) | 2023-01-13 |
| RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer | ✓ Link | 49.9 | 161 (T4) | RT-DETRv2-M | 2024-07-24 |
| YOLO-MS: Rethinking Multi-Scale Representation Learning for Real-time Object Detection | ✓ Link | 49.7 | 95 (RTX 3090) | YOLO-MS | 2023-08-10 |
| Octave-YOLO: Cross frequency detection network with octave convolution | | 49.7 | | Octave-YOLO-M | 2024-07-29 |
| RT-DETRv4: Painlessly Furthering Real-Time Object Detection with Vision Foundation Models | | 49.7 | 273 (T4) | RT-DETRv4-S | 2025-10-29 |
| RTMDet: An Empirical Study of Designing Real-Time Object Detectors | ✓ Link | 49.4 | 617 (RTX 3090) | RTMDet-m | 2022-12-14 |
| MHAF-YOLO: Multi-Branch Heterogeneous Auxiliary Fusion YOLO for accurate object detection | ✓ Link | 49.4 | | MHAF-YOLO-s (pretrained backbone) | 2025-02-07 |
| DAMO-YOLO : A Report on Real-Time Object Detection Design | ✓ Link | 49.2 | 233 (V100) | DAMO-YOLO-M | 2022-11-23 |
| Mamba YOLO: A Simple Baseline for Object Detection with State Space Model | ✓ Link | 49.1 | 455 (RTX 4090) | Mamba YOLO-B | 2024-06-09 |
| YOLO-Master: MOE-Accelerated with Specialized Transformers for Enhanced Real-time Detection | | 49.1 | | YOLO-Master-S | 2025-12-29 |
| DEIM: DETR with Improved Matching for Fast Convergence | ✓ Link | 49.0 | 287 (T4) | DEIM-D-FINE-S | 2024-12-05 |
| MHAF-YOLO: Multi-Branch Heterogeneous Auxiliary Fusion YOLO for accurate object detection | ✓ Link | 48.9 | | MHAF-YOLO-s | 2025-02-07 |
| RT-DETRv3: Real-time End-to-End Object Detection with Hierarchical Dense Positive Supervision | ✓ Link | 48.7 | 217 (T4) | RT-DETRv3-R18 | 2024-09-13 |
| Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models | ✓ Link | 48.6 | 400 (T4) | YOLO26s (Objects365 pre-training) | 2026-06-02 |
| D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement | ✓ Link | 48.5 | 287 (T4) | D-FINE-S | 2024-10-17 |
| HA-DETR: accelerating real-time object detection by replacing decoder self-attention | | 48.4 | 68 (V100) | HA-DETR (ResNet-18) | 2026-04-16 |
| LW-DETR: A Transformer Replacement to YOLO for Real-Time Detection | ✓ Link | 48.0 | 345 (T4) | LW-DETR-small (Objects365 pre-training) | 2024-06-05 |
| Hyper-YOLO: When Visual Object Detection Meets Hypergraph Computation | ✓ Link | 48.0 | 212 (T4) | Hyper-YOLO-S | 2024-08-09 |
| YOLOv12: Attention-Centric Real-Time Object Detectors | ✓ Link | 48.0 | 383 (T4) | YOLOv12-S | 2025-02-18 |
| YOLOv13: Real-Time Object Detection with Hypergraph-Enhanced Adaptive Visual Perception | ✓ Link | 48.0 | 336 (T4) | YOLOv13-S | 2025-06-21 |
| RF-DETR: Neural Architecture Search for Real-Time Detection Transformers | | 48.0 | 435 (T4) | RF-DETR-N (384px) (Objects365 pre-training) | 2025-11-12 |
| RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer | ✓ Link | 47.9 | 217 (T4) | RT-DETRv2-S | 2024-07-24 |
| Multi-Branch Auxiliary Fusion YOLO with Re-parameterization Heterogeneous Convolutional for accurate object detection | ✓ Link | 47.4 | | MAF-YOLO-s | 2024-07-05 |
| P$^2$HCT: Plug-and-Play Hierarchical C2F Transformer for Multi-Scale Feature Fusion | ✓ Link | 47.4 | | YOLOv11-P2HCT-S | 2025-05-19 |
| Ultralytics YOLO11 | ✓ Link | 47.0 | 400 (T4) | YOLO11s | 2024-09-27 |
| YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information | ✓ Link | 46.8 | | YOLOv9-S | 2024-02-21 |
| YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information | ✓ Link | 46.7 | | GELAN-S | 2024-02-21 |
| YOLO-PRO: Enhancing Instance-Specific Object Detection with Full-Channel Global Self-Attention | | 46.5 | 355 (T4) | YOLO-PRO-S (YOLOv8-based) | 2025-03-04 |
| XiYOLO: Energy-Aware Object Detection via Iterative Architecture Search and Scaling | | 46.5 | | XiYOLO-s | 2026-05-07 |
| YOLOv10: Real-Time End-to-End Object Detection | ✓ Link | 46.3 | 402 (T4) | YOLOv10-S (NMS-free) | 2024-05-23 |
| Gold-YOLO: Efficient Object Detector via Gather-and-Distribute Mechanism | ✓ Link | 46.1 | 286 (T4) | Gold-YOLO-S (self-distillation) | 2023-09-20 |
| DAMO-YOLO : A Report on Real-Time Object Detection Design | ✓ Link | 46.0 | 325 (V100) | DAMO-YOLO-S | 2022-11-23 |
| DEYO: DETR with YOLO for End-to-End Object Detection | ✓ Link | 45.8 | 299 (T4) | DEYO-S | 2024-02-26 |
| YOLO-MS: Rethinking Multi-Scale Representation Learning for Real-time Object Detection | ✓ Link | 45.4 | 137 (RTX 3090) | YOLO-MS-S | 2023-08-10 |
| YOLOv6 v3.0: A Full-Scale Reloading | ✓ Link | 45.0 | 339 (T4) | YOLOv6-S (self-distillation) | 2023-01-13 |
| Octave-YOLO: Cross frequency detection network with octave convolution | | 45.0 | | Octave-YOLO-S | 2024-07-29 |
| Ultralytics YOLOv8 | ✓ Link | 44.9 | 833 (A100) | YOLOv8s | 2023-01-10 |
| RTMDet: An Empirical Study of Designing Real-Time Object Detectors | ✓ Link | 44.6 | 820 (RTX 3090) | RTMDet-s | 2022-12-14 |
| Mamba YOLO: A Simple Baseline for Object Detection with State Space Model | ✓ Link | 44.5 | 667 (RTX 4090) | Mamba YOLO-T | 2024-06-09 |
| VajraV1 -- The most accurate Real Time Object Detector of the YOLO family | ✓ Link | 44.3 | 909 (RTX 4090) | VajraV1-Nano | 2025-12-15 |
| MHAF-YOLO: Multi-Branch Heterogeneous Auxiliary Fusion YOLO for accurate object detection | ✓ Link | 43.1 | | MHAF-YOLO-n (pretrained backbone) | 2025-02-07 |
| Real-Time Object Detection Meets DINOv3 | ✓ Link | 43.0 | | DEIMv2-Nano | 2025-09-25 |
| YOLO-MS: Rethinking Multi-Scale Representation Learning for Real-time Object Detection | ✓ Link | 42.8 | 141 (RTX 3090) | YOLO-MS-XS | 2023-08-10 |
| LW-DETR: A Transformer Replacement to YOLO for Real-Time Detection | ✓ Link | 42.6 | 500 (T4) | LW-DETR-tiny (Objects365 pre-training) | 2024-06-05 |
| P$^2$HCT: Plug-and-Play Hierarchical C2F Transformer for Multi-Scale Feature Fusion | ✓ Link | 42.6 | | YOLOv12-P2HCT-N | 2025-05-19 |
| Multi-Branch Auxiliary Fusion YOLO with Re-parameterization Heterogeneous Convolutional for accurate object detection | ✓ Link | 42.4 | | MAF-YOLO-n | 2024-07-05 |
| YOLO-Master: MOE-Accelerated with Specialized Transformers for Enhanced Real-time Detection | | 42.4 | | YOLO-Master-N | 2025-12-29 |
| MHAF-YOLO: Multi-Branch Heterogeneous Auxiliary Fusion YOLO for accurate object detection | ✓ Link | 42.3 | | MHAF-YOLO-n | 2025-02-07 |
| End-to-End Object Detection with Transformers | ✓ Link | 42.0 | 26 | Faster R-CNN-FPN+ (ResNet-50, reported by DETR) | 2020-05-26 |
| DAMO-YOLO : A Report on Real-Time Object Detection Design | ✓ Link | 42.0 | 397 (V100) | DAMO-YOLO-T | 2022-11-23 |
| Lightweight Backbone Networks Only Require Adaptive Lightweight Self-Attention Mechanisms | | 41.9 | 99 (RTX 4060 Ti) | LOLViT-X (YOLO head) | 2025-08-02 |
| Hyper-YOLO: When Visual Object Detection Meets Hypergraph Computation | ✓ Link | 41.8 | 364 (T4) | Hyper-YOLO-N | 2024-08-09 |
| YOLOv13: Real-Time Object Detection with Hypergraph-Enhanced Adaptive Visual Perception | ✓ Link | 41.6 | 508 (T4) | YOLOv13-N | 2025-06-21 |
| RTMDet: An Empirical Study of Designing Real-Time Object Detectors | ✓ Link | 41.1 | 1020 (RTX 3090) | RTMDet-tiny | 2022-12-14 |
| Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models | ✓ Link | 40.9 | 588 (T4) | YOLO26n (Objects365 pre-training) | 2026-06-02 |
| Rethinking Features-Fused-Pyramid-Neck for Object Detection | ✓ Link | 40.8 | | SYOLO-s | 2025-05-19 |
| YOLOv12: Attention-Centric Real-Time Object Detectors | ✓ Link | 40.6 | 610 (T4) | YOLOv12-N | 2025-02-18 |
| P$^2$HCT: Plug-and-Play Hierarchical C2F Transformer for Multi-Scale Feature Fusion | ✓ Link | 40.3 | | YOLOv11-P2HCT-N | 2025-05-19 |
| Gold-YOLO: Efficient Object Detector via Gather-and-Distribute Mechanism | ✓ Link | 39.9 | 563 (T4) | Gold-YOLO-N (self-distillation) | 2023-09-20 |
| DEYO: DETR with YOLO for End-to-End Object Detection | ✓ Link | 39.7 | 396 (T4) | DEYO-N | 2024-02-26 |
| Ultralytics YOLO11 | ✓ Link | 39.5 | 667 (T4) | YOLO11n | 2024-09-27 |
| YOLOv10: Real-Time End-to-End Object Detection | ✓ Link | 38.5 | 543 (T4) | YOLOv10-N (NMS-free) | 2024-05-23 |
| Hyper-YOLO: When Visual Object Detection Meets Hypergraph Computation | ✓ Link | 38.5 | 404 (T4) | Hyper-YOLO-T | 2024-08-09 |
| Real-Time Object Detection Meets DINOv3 | ✓ Link | 38.5 | | DEIMv2-Pico | 2025-09-25 |
| YOLO-PRO: Enhancing Instance-Specific Object Detection with Full-Channel Global Self-Attention | | 38.3 | 654 (T4) | YOLO-PRO-N (YOLOv8-based) | 2025-03-04 |
| Lightweight Backbone Networks Only Require Adaptive Lightweight Self-Attention Mechanisms | | 37.8 | 113 (RTX 4060 Ti) | LOLViT-S (YOLO head) | 2025-08-02 |
| DEYO: DETR with YOLO for End-to-End Object Detection | ✓ Link | 37.6 | 497 (T4) | DEYO-tiny | 2024-02-26 |
| HA-DETR: accelerating real-time object detection by replacing decoder self-attention | | 37.6 | 97 (V100) | HA-DETR (MobileNetV3-small) | 2026-04-16 |
| YOLOv6 v3.0: A Full-Scale Reloading | ✓ Link | 37.5 | 779 (T4) | YOLOv6-N (self-distillation) | 2023-01-13 |
| Octave-YOLO: Cross frequency detection network with octave convolution | | 37.5 | 230 (RTX 3090) | Octave-YOLO-N | 2024-07-29 |
| Ultralytics YOLOv8 | ✓ Link | 37.3 | 1010 (A100) | YOLOv8n | 2023-01-10 |
| Rethinking Features-Fused-Pyramid-Neck for Object Detection | ✓ Link | 35.9 | | SYOLO-n (416px) | 2025-05-19 |
| Ultralytics YOLOv5 | ✓ Link | 28.0 | 159 (V100) | YOLOv5n (official repo) | 2021-10-12 |
| 3A-YOLO: New Real-Time Object Detectors with Triple Discriminative Awareness and Coordinated Representations | | 25.9 | | 3A-YOLO-Tiny | 2024-12-10 |
| 3A-YOLO: New Real-Time Object Detectors with Triple Discriminative Awareness and Coordinated Representations | | 23.5 | | 3A-YOLO-Nano | 2024-12-10 |