Crafting GBD-Net for Object Detection

The visual cues from multiple support regions of different sizes and resolutions are complementary in classifying a candidate box in object detection. Effective integration of local and contextual visual cues from these regions has become a fundamental problem in object detection. In this paper, we propose a gated bi-directional CNN (GBD-Net) to pass messages among features from different support regions during both feature learning and feature extraction. Such message passing can be implemented through convolution between neighboring support regions in two directions and can be conducted in various layers. Therefore, local and contextual visual patterns can validate the existence of each other by learning their nonlinear relationships and their close interactions are modeled in a more complex way. It is also shown that message passing is not always helpful but dependent on individual samples. Gated functions are therefore needed to control message transmission, whose on-or-offs are controlled by extra visual evidence from the input sample. The effectiveness of GBD-Net is shown through experiments on three object detection datasets, ImageNet, Pascal VOC2007 and Microsoft COCO. Besides the GBD-Net, this paper also shows the details of our approach in winning the ImageNet object detection challenge of 2016, with source code provided on https://github.com/craftGBD/craftGBD. In this winning system, the modified GBD-Net, new pretraining scheme and better region proposal designs are provided. We also show the effectiveness of different network structures and existing techniques for object detection, such as multi-scale testing, left-right flip, bounding box voting, NMS, and context.

• Publication year

  2016

• Venue

  IEEE Transactions on Pattern Analysis and Machine Intelligence

• Publication date

  2016-10-08

• Fields of study

  Medicine, Computer Science

• Identifiers
  DOI 10.1109/TPAMI.2017.2745563arXiv 1610.02579PMID 28858785
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar, PubMed

• No claims are published for this paper.

• No concepts are published for this paper.

• Jointly Learning Deep Features, Deformable Parts, Occlusion and Classification for Pedestrian Detection

  2018cited by this paper
• Chained Cascade Network for Object Detection

  2017cited by this paper
• Zoom Out-and-In Network with Recursive Training for Object Proposal

  2017cited by this paper
• CRAFT Objects from Images

  2016influential reference
• PolyNet: A Pursuit of Structural Diversity in Very Deep Networks

  2016cited by this paper
• Gated Bi-directional CNN for Object Detection

  2016influential reference
• Multi-Bias Non-linear Activation in Deep Neural Networks

  2016cited by this paper
• Identity Mappings in Deep Residual Networks

  2016influential reference
• Deep Networks with Stochastic Depth

  2016influential reference
• A MultiPath Network for Object Detection

  2016cited by this paper
• Factors in Finetuning Deep Model for Object Detection with Long-Tail Distribution

  2016cited by this paper
• Attend Refine Repeat: Active Box Proposal Generation via In-Out Localization

  2016cited by this paper
• Structured Feature Learning for Pose Estimation

  2016cited by this paper
• Window-Object Relationship Guided Representation Learning for Generic Object Detections

  2015cited by this paper
• You Only Look Once: Unified, Real-Time Object Detection

  2015cited by this paper
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

  2015influential reference
• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks

  2015influential reference
• LocNet: Improving Localization Accuracy for Object Detection

  2015influential reference
• SSD: Single Shot MultiBox Detector

  2015cited by this paper
• Deep Residual Learning for Image Recognition

  2015influential reference
• Inside-Outside Net: Detecting Objects in Context with Skip Pooling and Recurrent Neural Networks

  2015influential reference
• Object detection by labeling superpixels

  2015cited by this paper
• Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification

  2015cited by this paper
• Object Detection via a Multi-region and Semantic Segmentation-Aware CNN Model

  2015influential reference
• What Makes for Effective Detection Proposals?

  2015cited by this paper
• Rethinking the Inception Architecture for Computer Vision

  2015cited by this paper
• Learning Deep Representation with Large-Scale Attributes

  2015cited by this paper
• Boosting Object Proposals: From Pascal to COCO

  2015cited by this paper
• Fast R-CNN

  2015influential reference
• Improving object detection with deep convolutional networks via Bayesian optimization and structured prediction

  2015cited by this paper
• Scalable, High-Quality Object Detection

  2014cited by this paper
• ImageNet Large Scale Visual Recognition Challenge

  2014cited by this paper
• Multiscale Combinatorial Grouping

  2014cited by this paper
• Edge Boxes: Locating Object Proposals from Edges

  2014cited by this paper
• DeepID-Net: Deformable deep convolutional neural networks for object detection

  2014influential reference
• Deformable part models are convolutional neural networks

  2014cited by this paper
• Microsoft COCO: Common Objects in Context

  2014influential reference
• Very Deep Convolutional Networks for Large-Scale Image Recognition

  2014influential reference
• Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition

  2014cited by this paper
• Going deeper with convolutions

  2014influential reference
• Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation

  2013influential reference
• Learning Hierarchical Features for Scene Labeling

  2013cited by this paper
• OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks

  2013cited by this paper
• ImageNet classification with deep convolutional neural networks

  2012cited by this paper
• Segmentation as selective search for object recognition

  2011cited by this paper
• Understanding the difficulty of training deep feedforward neural networks

  2010cited by this paper
• The Pascal Visual Object Classes (VOC) Challenge

  2010cited by this paper
• Gradient-based learning applied to document recognition

  1998cited by this paper
• Long Short-Term Memory

  1997cited by this paper

• An Integrated Deep Learning and LLM Model for Burn Wound Depth Recognition.

  2025cites this paper
• SCD-SG Net: a small object detection method for unmanned aerial vehicles based on spatial and channel details with semantic guidance

  2025cites this paper
• Multi-Scale Hierarchical Feature Fusion for Infrared Small-Target Detection

  2025cites this paper
• Advancements in Obstacle Intrusion Detection Methods for Rail Transit: A Comprehensive Review

  2025cites this paper
• Dissecting of the two-stages object detection models architecture and performance

  2024cites this paper
• A review of small object detection based on deep learning

  2024cites this paper
• A Comprehensive Review of Convolutional Neural Networks for Defect Detection in Industrial Applications

  2024cites this paper
• Dynamic Malware Analysis and Node Behavior Monitoring with Self-Replication for Enhanced Network Intrusion Detection

  2024cites this paper
• Composite Backbone Small Object Detection Based on Context and Multi-Scale Information with Attention Mechanism

  2024cites this paper
• Discretion-Preserving with Data Mining Drive Distribution Scheme with a Universal Social Grid Web for Vans Using Vast Data

  2023cites this paper
• Deep Learning object detection models: evolution and evaluation

  2023cites this paper
• A Task-Balanced Multiscale Adaptive Fusion Network for Object Detection in Remote Sensing Images

  2023cites this paper
• GLFFNet: A Global and Local Features Fusion Network with Biencoder for Remote Sensing Image Segmentation

  2023cites this paper
• Augmented weighted bidirectional feature pyramid network for marine object detection

  2023cites this paper
• Dynamic Local and Global Context Exploration for Small Object Detection

  2023cites this paper
• Latent knowledge reasoning incorporated for multi-fitting decoupling detection on electric transmission line

  2023cites this paper
• Crowd counting from single images using recursive multi-pathway zooming and foreground enhancement

  2023cites this paper
• Deep CNN-based visual defect detection: Survey of current literature

  2023cites this paper
• Ship detection with deep learning: a survey

  2023cites this paper
• Fighting Money Laundering with Statistics and Machine Learning: An Introduction and Review

  2022cites this paper
• Study on the Lightweighting Strategy of Target Detection Model with Deep Learning

  2022cites this paper
• Real-Time Oil Leakage Detection on Aftermarket Motorcycle Damping System with Convolutional Neural Networks

  2022cites this paper
• Transformer-Based Context Condensation for Boosting Feature Pyramids in Object Detection

  2022cites this paper
• Exploring Contextual Relationships for Cervical Abnormal Cell Detection

  2022cites this paper
• Multi-Scale Feature Fusion Convolutional Neural Network for Indoor Small Target Detection

  2022cites this paper
• Induced Pluripotent Stem Cell-Based Drug Screening by Use of Artificial Intelligence

  2022cites this paper
• Evaluation Methodology for Object Detection and Tracking in Bounding Box Based Perception Modules

  2022cites this paper
• Discriminative information restoration and extraction for weakly supervised low-resolution fine-grained image recognition

  2022cites this paper
• Fighting Money Laundering With Statistics and Machine Learning

  2022cites this paper
• DDBN: Dual detection branch network for semantic diversity predictions

  2022cites this paper
• An Analysis on Object Recognition Using Convolutional Neural Networks

  2021cites this paper
• TCANet: Triple Context-Aware Network for Weakly Supervised Object Detection in Remote Sensing Images

  2021cites this paper
• MSB R-CNN: A Multi-Stage Balanced Defect Detection Network

  2021cites this paper
• Context augmentation for object detection

  2021cites this paper
• Flower Detection Using Object Analysis: New Ways to Quantify Plant Phenology in a Warming Tundra Biome

  2021cites this paper
• Artificial intelligence in the management and treatment of burns: a systematic review

  2021cites this paper
• Exploiting Full-Scale Feature for Remote Sensing Object Detection Based on Refined Feature Mining and Adaptive Fusion

  2021cites this paper
• Towards Balanced Learning for Instance Recognition

  2021cites this paper
• Object detection algorithm based on feature enhancement

  2021cites this paper
• ABNet: Adaptive Balanced Network for Multi-scale Object Detection in Remote Sensing Imagery

  2021cites this paper
• FIN

  2020cites this paper
• Progressive Contextual Instance Refinement for Weakly Supervised Object Detection in Remote Sensing Images

  2020cites this paper
• Syncretic-NMS: A Merging Non-Maximum Suppression Algorithm for Instance Segmentation

  2020cites this paper
• Auxiliary Detection Head for One-Stage Object Detection

  2020cites this paper
• Diverse receptive field network with context aggregation for fast object detection

  2020cites this paper
• KPNet: Towards Minimal Face Detector

  2020cites this paper
• Revisiting the Sibling Head in Object Detector

  2020cites this paper
• 1st Place Solutions for OpenImage2019 - Object Detection and Instance Segmentation

  2020cites this paper
• Recursive Context Routing for Object Detection

  2020cites this paper
• A survey on instance segmentation: state of the art

  2020cites this paper
• A Contextual Bidirectional Enhancement Method for Remote Sensing Image Object Detection

  2020influential citation
• Multi-Scale Based Context-Aware Net for Action Detection

  2020influential citation
• Shape and Texture Aware Graph Processing

  2020cites this paper
• Deep Learning Approaches for Object Detection

  2020cites this paper
• Probabilistic Graph Attention Network With Conditional Kernels for Pixel-Wise Prediction

  2020cites this paper
• Stem cell imaging through convolutional neural networks: current issues and future directions in artificial intelligence technology

  2020cites this paper
• Multi-field of View Aggregation and Context Encoding for Single-Stage Nucleus Recognition

  2020cites this paper
• Beyond Short-Term Snippet: Video Relation Detection With Spatio-Temporal Global Context

  2020cites this paper
• Renovating Parsing R-CNN for Accurate Multiple Human Parsing

  2020cites this paper
• RoI Feature Propagation for Video Object Detection

  2020cites this paper
• Spatial attention model based target detection for aerial robotic systems

  2019cites this paper
• Object Detection based on Region Decomposition and Assembly

  2019cites this paper
• Gated bidirectional feature pyramid network for accurate one-shot detection

  2019cites this paper
• Visual Semantic Information Pursuit: A Survey

  2019cites this paper
• Accurate Monocular 3D Object Detection via Color-Embedded 3D Reconstruction for Autonomous Driving

  2019cites this paper
• Libra R-CNN: Towards Balanced Learning for Object Detection

  2019cites this paper
• Deep learning - Method overview and review of use for fruit detection and yield estimation

  2019cites this paper
• Object Detection in 20 Years: A Survey

  2019cites this paper
• Monocular Depth Estimation with Directional Consistency by Deep Networks

  2019cites this paper
• Feature Pyramid Reconfiguration With Consistent Loss for Object Detection

  2019cites this paper
• Detecting small faces in the wild based on generative adversarial network and contextual information

  2019cites this paper
• Synthetical application of multi-feature map detection and multi-branch convolution

  2019cites this paper
• A Feature Aggregation Convolutional Neural Network for Remote Sensing Scene Classification

  2019influential citation
• The application of convolutional neural network to stem cell biology

  2019cites this paper
• Ship Detection Using Machine Learning and Optical Imagery in the Maritime Environment

  2019cites this paper
• Scale adaptive image cropping for UAV object detection

  2019cites this paper
• DFPENet-geology: A Deep Learning Framework for High Precision Recognition and Segmentation of Co-seismic Landslides

  2019cites this paper
• HiCoRe: Visual Hierarchical Context-Reasoning

  2019cites this paper
• Object Detection and Classification using Convolutional Neural Network

  2019cites this paper
• Deepsub: A Novel Subset Selection Framework for Training Deep Learning Architectures

  2019cites this paper
• ASSD: Attentive Single Shot Multibox Detector

  2019cites this paper
• Probabilistic Online Learning of Appearance and Structure for Robotics

  2019cites this paper
• Comprehensive Review of Artificial Neural Network Applications to Pattern Recognition

  2019cites this paper
• A Context Aware Deep Learning Architecture for Object Detection

  2019cites this paper
• Multi-scale Feature and Spatial Relation Inference for Object Detection

  2019cites this paper
• AugFPN: Improving Multi-Scale Feature Learning for Object Detection

  2019cites this paper
• Taking a Look at Small-Scale Pedestrians and Occluded Pedestrians

  2019cites this paper
• Robotic grasp detection based on image processing and random forest

  2019cites this paper
• Apprentissage neuronal profond pour l'analyse de contenus multimodaux et temporels. (Deep learning for multimodal and temporal contents analysis)

  2019cites this paper
• Dense pose estimation of deformable objects

  2019cites this paper
• A hierarchical contextual attention-based network for sequential recommendation

  2019cites this paper
• Effective and annotation efficient deep learning for image understanding

  2018cites this paper
• Towards the Design of Neural Network Framework for Object Recognition and Target Region Refining for Smart Transportation Systems

  2018influential citation
• An Application of Transfer Learning for Maritime Vision Processing Using Machine Learning

  2018cites this paper
• DeepWeeds: A Multiclass Weed Species Image Dataset for Deep Learning

  2018cites this paper
• Context Refinement for Object Detection

  2018cites this paper
• Evolutionary Deep Learning: A Genetic Programming Approach to Image Classification

  2018cites this paper
• An Efficient Object Detection Algorithm Based on Compressed Networks

  2018cites this paper
• Multi-scale Context Intertwining for Semantic Segmentation

  2018cites this paper
• Key Parts Context and Scene Geometry in Human Head Detection

  2018cites this paper