Soft Sampling for Robust Object Detection

Abstract

We study the robustness of object detection under the presence of missing annotations. In this setting, the unlabeled object instances will be treated as background, which will generate an incorrect training signal for the detector. Interestingly, we observe that after dropping 30% of the annotations (and labeling them as background), the performance of CNN-based object detectors like Faster-RCNN only drops by 5% on the PASCAL VOC dataset. We provide a detailed explanation for this result. To further bridge the performance gap, we propose a simple yet effective solution, called Soft Sampling. Soft Sampling re-weights the gradients of RoIs as a function of overlap with positive instances. This ensures that the uncertain background regions are given a smaller weight compared to the hardnegatives. Extensive experiments on curated PASCAL VOC datasets demonstrate the effectiveness of the proposed Soft Sampling method at different annotation drop rates. Finally, we show that on OpenImagesV3, which is a real-world dataset with missing annotations, Soft Sampling outperforms standard detection baselines by over 3%.

Accepted in BMVC 2019

References in corpus (7)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• R-FCN: Object Detection via Region-based Fully Convolutional Networks
• The Open Images Dataset V4: Unified image classification, object detection, and visual relationship detection at scale
• Deformable Convolutional Networks
• MegDet: A Large Mini-Batch Object Detector
• Solution for Large-Scale Hierarchical Object Detection Datasets with Incomplete Annotation and Data Imbalance

Cited by in corpus (16)

• Object Detection as a Positive-Unlabeled Problem
• Solution for Large-Scale Hierarchical Object Detection Datasets with Incomplete Annotation and Data Imbalance
• OMNIA Faster R-CNN: Detection in the wild through dataset merging and soft distillation
• Active learning using weakly supervised signals for quality inspection
• Deep Regionlets for Object Detection
• Faster ILOD: Incremental Learning for Object Detectors based on Faster RCNN
• Co-mining: Self-Supervised Learning for Sparsely Annotated Object Detection
• Data-Uncertainty Guided Multi-Phase Learning for Semi-Supervised Object Detection
• Large-Scale Object Detection in the Wild from Imbalanced Multi-Labels
• An End-to-End Framework For Universal Lesion Detection With Missing Annotations
• Self-supervised Robust Object Detectors from Partially Labelled Datasets
• Detecting 11K Classes: Large Scale Object Detection without Fine-Grained Bounding Boxes
• Dynamic Supervisor for Cross-dataset Object Detection
• Unsupervised Object Detection with LiDAR Clues
• Decoupled Gradient Harmonized Detector for Partial Annotation: Application to Signet Ring Cell Detection
• TagMe: GPS-Assisted Automatic Object Annotation in Videos