Weighted boxes fusion: Ensembling boxes from different object detection models

Abstract

In this work, we present a novel method for combining predictions of object detection models: weighted boxes fusion. Our algorithm utilizes confidence scores of all proposed bounding boxes to constructs the averaged boxes. We tested method on several datasets and evaluated it in the context of the Open Images and COCO Object Detection tracks, achieving top results in these challenges. The source code is publicly available at https://github.com/ZFTurbo/Weighted-Boxes-Fusion

References in corpus (7)

• CheXNet: Radiologist-Level Pneumonia Detection on Chest X-Rays with Deep Learning
• The Open Images Dataset V4: Unified image classification, object detection, and visual relationship detection at scale
• DetectoRS: Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolution
• Deep Learning for Automatic Pneumonia Detection
• Daedalus: Breaking Non-Maximum Suppression in Object Detection via Adversarial Examples
• 1st Place Solution for Waymo Open Dataset Challenge -- 3D Detection and Domain Adaptation
• 2nd Place Solution for Waymo Open Dataset Challenge -- 2D Object Detection

Cited by in corpus (28)

• Fracture Detection in Wrist X-ray Images Using Deep Learning-Based Object Detection Models
• TPH-YOLOv5: Improved YOLOv5 Based on Transformer Prediction Head for Object Detection on Drone-captured Scenarios
• An original framework for Wheat Head Detection using Deep, Semi-supervised and Ensemble Learning within Global Wheat Head Detection (GWHD) Dataset
• Towards Better Accuracy-efficiency Trade-offs: Divide and Co-training
• Deep learning in computed tomography pulmonary angiography imaging: a dual-pronged approach for pulmonary embolism detection
• MVA2023 Small Object Detection Challenge for Spotting Birds: Dataset, Methods, and Results
• Optimizing YOLOv7 for Semiconductor Defect Detection
• 1st Place Solution for Waymo Open Dataset Challenge -- 3D Detection and Domain Adaptation
• Cross-domain Federated Object Detection
• YOLOv8 for Defect Inspection of Hexagonal Directed Self-Assembly Patterns: A Data-Centric Approach
• Object Detection in Thermal Images Using Deep Learning for Unmanned Aerial Vehicles
• VIPriors 1: Visual Inductive Priors for Data-Efficient Deep Learning Challenges
• Deep Learning in Diabetic Foot Ulcers Detection: A Comprehensive Evaluation
• Cycle Self-Training for Semi-Supervised Object Detection with Distribution Consistency Reweighting
• GIAOTracker: A comprehensive framework for MCMOT with global information and optimizing strategies in VisDrone 2021
• Collective PV-RCNN: A Novel Fusion Technique using Collective Detections for Enhanced Local LiDAR-Based Perception
• 2nd Place Solution for Waymo Open Dataset Challenge -- 2D Object Detection
• 1st Place Solution for ICDAR 2021 Competition on Mathematical Formula Detection
• ICDAR 2023 Competition on Robust Layout Segmentation in Corporate Documents
• Data-Uncertainty Guided Multi-Phase Learning for Semi-Supervised Object Detection
• Global Wheat Challenge 2020: Analysis of the competition design and winning models
• Boosting ship detection in SAR images with complementary pretraining techniques
• 1st Place Solutions for UG2+ Challenge 2021 -- (Semi-)supervised Face detection in the low light condition
• An efficient and flexible inference system for serving heterogeneous ensembles of deep neural networks
• Automating the Wildfire Detection and Scheduling Pipeline with Maneuverable Earth Observation Satellites
• Two-step Domain Adaptation for Mitosis Cell Detection in Histopathology Images
• Model Composition: Can Multiple Neural Networks Be Combined into a Single Network Using Only Unlabeled Data?
• Towards Comprehensive Monocular Depth Estimation: Multiple Heads Are Better Than One