Crafting GBD-Net for Object Detection

Abstract

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. This paper also shows the details of our approach in wining the ImageNet object detection challenge of 2016, with source code provided on \url{https://github.com/craftGBD/craftGBD}.

This paper shows the details of our approach in wining the ImageNet object detection challenge of 2016, with source code provided on \url{https://github.com/craftGBD/craftGBD}. The preliminary version of this paper is presented at the ECCV. Xingyu Zeng and Wanli Ouyang contributed equally

References in corpus (10)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
• What makes for effective detection proposals?
• Scalable, High-Quality Object Detection
• Deep Networks with Stochastic Depth
• DeepID-Net: Deformable Deep Convolutional Neural Networks for Object Detection
• Deformable Part Models are Convolutional Neural Networks
• Multi-Bias Non-linear Activation in Deep Neural Networks

Cited by in corpus (11)

• Path Aggregation Network for Instance Segmentation
• Learning Deep Structured Multi-Scale Features using Attention-Gated CRFs for Contour Prediction
• Deep Continuous Conditional Random Fields with Asymmetric Inter-object Constraints for Online Multi-object Tracking
• Learning Deep Context-aware Features over Body and Latent Parts for Person Re-identification
• A Survey on Deep Learning Methods for Robot Vision
• Integrated Object Detection and Tracking with Tracklet-Conditioned Detection
• Optical Flow Guided Feature: A Fast and Robust Motion Representation for Video Action Recognition
• A Hierarchical Contextual Attention-based GRU Network for Sequential Recommendation
• Recurrent Scale Approximation for Object Detection in CNN
• KPNet: Towards Minimal Face Detector
• Monocular Depth Estimation with Directional Consistency by Deep Networks