RepPoints: Point Set Representation for Object Detection

Abstract

Modern object detectors rely heavily on rectangular bounding boxes, such as anchors, proposals and the final predictions, to represent objects at various recognition stages. The bounding box is convenient to use but provides only a coarse localization of objects and leads to a correspondingly coarse extraction of object features. In this paper, we present \textbf{RepPoints} (representative points), a new finer representation of objects as a set of sample points useful for both localization and recognition. Given ground truth localization and recognition targets for training, RepPoints learn to automatically arrange themselves in a manner that bounds the spatial extent of an object and indicates semantically significant local areas. They furthermore do not require the use of anchors to sample a space of bounding boxes. We show that an anchor-free object detector based on RepPoints can be as effective as the state-of-the-art anchor-based detection methods, with 46.5 AP and 67.4 $A{P}_{50}$ on the COCO test-dev detection benchmark, using ResNet-101 model. Code is available at https://github.com/microsoft/RepPoints.

arXiv commentInternational Conference on Computer Vision (ICCV), 2019

References in corpus (14)

• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• YOLOv3: An Incremental Improvement
• R-FCN: Object Detection via Region-based Fully Convolutional Networks
• DeepPose: Human Pose Estimation via Deep Neural Networks
• DSSD : Deconvolutional Single Shot Detector
• Learning to Segment Object Candidates
• Associative Embedding: End-to-End Learning for Joint Detection and Grouping
• DenseBox: Unifying Landmark Localization with End to End Object Detection
• The Open Images Dataset V4: Unified image classification, object detection, and visual relationship detection at scale
• Light-Head R-CNN: In Defense of Two-Stage Object Detector
• MetaAnchor: Learning to Detect Objects with Customized Anchors
• Consistent Optimization for Single-Shot Object Detection
• Anchor Box Optimization for Object Detection
• DeNet: Scalable Real-time Object Detection with Directed Sparse Sampling

Cited by in corpus (16)

• Generalized Focal Loss: Learning Qualified and Distributed Bounding Boxes for Dense Object Detection
• AutoAssign: Differentiable Label Assignment for Dense Object Detection
• DeepSOCIAL: Social Distancing Monitoring and Infection Risk Assessment in COVID-19 Pandemic
• G-Rep: Gaussian Representation for Arbitrary-Oriented Object Detection
• Revisiting Feature Alignment for One-stage Object Detection
• RelationNet++: Bridging Visual Representations for Object Detection via Transformer Decoder
• Occluded Prohibited Items Detection: an X-ray Security Inspection Benchmark and De-occlusion Attention Module
• Guided Attention Network for Object Detection and Counting on Drones
• Revisiting the Sibling Head in Object Detector
• Learning from Noisy Anchors for One-stage Object Detection
• Decoupled and Memory-Reinforced Networks: Towards Effective Feature Learning for One-Step Person Search
• Differentiable Multi-Granularity Human Representation Learning for Instance-Aware Human Semantic Parsing
• Dually Supervised Feature Pyramid for Object Detection and Segmentation
• Scope Head for Accurate Localization in Object Detection
• Bootstrap Your Object Detector via Mixed Training
• DSIC: Dynamic Sample-Individualized Connector for Multi-Scale Object Detection