CornerNet: Detecting Objects as Paired Keypoints
arXiv:1808.01244
Abstract
We propose CornerNet, a new approach to object detection where we detect an object bounding box as a pair of keypoints, the top-left corner and the bottom-right corner, using a single convolution neural network. By detecting objects as paired keypoints, we eliminate the need for designing a set of anchor boxes commonly used in prior single-stage detectors. In addition to our novel formulation, we introduce corner pooling, a new type of pooling layer that helps the network better localize corners. Experiments show that CornerNet achieves a 42.2% AP on MS COCO, outperforming all existing one-stage detectors.
arXiv commentExtended version with additional results. Test AP on MS COOO improved from 42.1% to 42.2% after a bug fix
References in corpus (23)
- Adam: A Method for Stochastic Optimization
- 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
- DSSD : Deconvolutional Single Shot Detector
- Focal Loss for Dense Object Detection
- Associative Embedding: End-to-End Learning for Joint Detection and Grouping
- Feature Pyramid Networks for Object Detection
- Deformable Convolutional Networks
- Cascade R-CNN: Delving into High Quality Object Detection
- Beyond Skip Connections: Top-Down Modulation for Object Detection
- Light-Head R-CNN: In Defense of Two-Stage Object Detector
- Pixels to Graphs by Associative Embedding
- Single-Shot Refinement Neural Network for Object Detection
- RON: Reverse Connection with Objectness Prior Networks for Object Detection
- Dual Path Networks
- Point Linking Network for Object Detection
- Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection
- An Analysis of Scale Invariance in Object Detection - SNIP
- DeNet: Scalable Real-time Object Detection with Directed Sparse Sampling
- Deep Regionlets for Object Detection
- Feature Selective Networks for Object Detection
Cited by in corpus (27)
- Deep Learning for UAV-based Object Detection and Tracking: A Survey
- Generalized Focal Loss: Learning Qualified and Distributed Bounding Boxes for Dense Object Detection
- AutoAssign: Differentiable Label Assignment for Dense Object Detection
- Deep learning for detection and segmentation of artefact and disease instances in gastrointestinal endoscopy
- RAMP-CNN: A Novel Neural Network for Enhanced Automotive Radar Object Recognition
- Review of data analysis in vision inspection of power lines with an in-depth discussion of deep learning technology
- SkyNet: a Hardware-Efficient Method for Object Detection and Tracking on Embedded Systems
- Revisiting Feature Alignment for One-stage Object Detection
- SiamFC++: Towards Robust and Accurate Visual Tracking with Target Estimation Guidelines
- Decoupled Classification Refinement: Hard False Positive Suppression for Object Detection
- FoodLogoDet-1500: A Dataset for Large-Scale Food Logo Detection via Multi-Scale Feature Decoupling Network
- HRDNet: High-resolution Detection Network for Small Objects
- M2Det: A Single-Shot Object Detector based on Multi-Level Feature Pyramid Network
- Occluded Prohibited Items Detection: an X-ray Security Inspection Benchmark and De-occlusion Attention Module
- AFP-Net: Realtime Anchor-Free Polyp Detection in Colonoscopy
- Reducing Label Noise in Anchor-Free Object Detection
- Decoupled IoU Regression for Object Detection
- Unstructured Road Vanishing Point Detection Using the Convolutional Neural Network and Heatmap Regression
- Box-level Segmentation Supervised Deep Neural Networks for Accurate and Real-time Multispectral Pedestrian Detection
- Visible Feature Guidance for Crowd Pedestrian Detection
- Inter-Homines: Distance-Based Risk Estimation for Human Safety
- KPNet: Towards Minimal Face Detector
- Pixel-Semantic Revise of Position Learning A One-Stage Object Detector with A Shared Encoder-Decoder
- An End-to-End Autofocus Camera for Iris on the Move
- DSIC: Dynamic Sample-Individualized Connector for Multi-Scale Object Detection
- A Light-Weight Object Detection Framework with FPA Module for Optical Remote Sensing Imagery
- Perceiving Traffic from Aerial Images