Probabilistic two-stage detection

Abstract

We develop a probabilistic interpretation of two-stage object detection. We show that this probabilistic interpretation motivates a number of common empirical training practices. It also suggests changes to two-stage detection pipelines. Specifically, the first stage should infer proper object-vs-background likelihoods, which should then inform the overall score of the detector. A standard region proposal network (RPN) cannot infer this likelihood sufficiently well, but many one-stage detectors can. We show how to build a probabilistic two-stage detector from any state-of-the-art one-stage detector. The resulting detectors are faster and more accurate than both their one- and two-stage precursors. Our detector achieves 56.4 mAP on COCO test-dev with single-scale testing, outperforming all published results. Using a lightweight backbone, our detector achieves 49.2 mAP on COCO at 33 fps on a Titan Xp, outperforming the popular YOLOv4 model.

Code is available at https://github.com/xingyizhou/CenterNet2

References in corpus (20)

• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• YOLOv4: Optimal Speed and Accuracy of Object Detection
• YOLOv3: An Incremental Improvement
• Res2Net: A New Multi-scale Backbone Architecture
• Deformable DETR: Deformable Transformers for End-to-End Object Detection
• Focal Loss for Dense Object Detection
• MMDetection: Open MMLab Detection Toolbox and Benchmark
• The Open Images Dataset V4: Unified image classification, object detection, and visual relationship detection at scale
• FreeAnchor: Learning to Match Anchors for Visual Object Detection
• AutoAssign: Differentiable Label Assignment for Dense Object Detection
• Hybrid Task Cascade for Instance Segmentation
• DetectoRS: Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolution
• Scaled-YOLOv4: Scaling Cross Stage Partial Network
• Revisiting Feature Alignment for One-stage Object Detection
• Region Proposal by Guided Anchoring
• Cascade RPN: Delving into High-Quality Region Proposal Network with Adaptive Convolution
• FCOS: A simple and strong anchor-free object detector
• CentripetalNet: Pursuing High-quality Keypoint Pairs for Object Detection
• Corner Proposal Network for Anchor-free, Two-stage Object Detection
• Dense RepPoints: Representing Visual Objects with Dense Point Sets

Cited by in corpus (8)

• You Only Learn One Representation: Unified Network for Multiple Tasks
• Focal Self-attention for Local-Global Interactions in Vision Transformers
• TPH-YOLOv5: Improved YOLOv5 Based on Transformer Prediction Head for Object Detection on Drone-captured Scenarios
• Dynamic Head: Unifying Object Detection Heads with Attentions
• PETDet: Proposal Enhancement for Two-Stage Fine-Grained Object Detection
• Challenges for Monocular 6D Object Pose Estimation in Robotics
• FBNetV5: Neural Architecture Search for Multiple Tasks in One Run
• Progressive Hard-case Mining across Pyramid Levels for Object Detection