D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement
Preprint
Abstract
We introduce D-FINE, a powerful real-time object detector that achieves outstanding localization precision by redefining the bounding box regression task in DETR models. D-FINE comprises two key components: Fine-grained Distribution Refinement (FDR) and Global Optimal Localization Self-Distillation (GO-LSD). FDR transforms the regression process from predicting fixed coordinates to iteratively refining probability distributions, providing a fine-grained intermediate representation that significantly enhances localization accuracy. GO-LSD is a bidirectional optimization strategy that transfers localization knowledge from refined distributions to shallower layers through self-distillation, while also simplifying the residual prediction tasks for deeper layers. Additionally, D-FINE incorporates lightweight optimizations in computationally intensive modules and operations, achieving a better balance between speed and accuracy. Specifically, D-FINE-L / X achieves 54.0% / 55.8% AP on the COCO dataset at 124 / 78 FPS on an NVIDIA T4 GPU. When pretrained on Objects365, D-FINE-L / X attains 57.1% / 59.3% AP, surpassing all existing real-time detectors. Furthermore, our method significantly enhances the performance of a wide range of DETR models by up to 5.3% AP with negligible extra parameters and training costs. Our code and pretrained models: https://github.com/Peterande/D-FINE.Keywords
All Related Versions
This publication has 13 references indexed in Scilit:
- YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information2024
- Generalized Focal Loss V2: Learning Reliable Localization Quality Estimation for Dense Object Detection2021
- General Instance Distillation for Object Detection2021
- Distilling Object Detectors via Decoupled Features2021
- Offset Bin Classification Network for Accurate Object Detection2020
- Gaussian YOLOv3: An Accurate and Fast Object Detector Using Localization Uncertainty for Autonomous Driving2019
- Distilling Object Detectors With Fine-Grained Feature Imitation2019
- Mimicking Very Efficient Network for Object Detection2017
- You Only Look Once: Unified, Real-Time Object Detection2016
- Microsoft COCO: Common Objects in Context2014