Prime Sample Attention in Object Detection

Abstract

It is a common paradigm in object detection frameworks to treat all samples equally and target at maximizing the performance on average. In this work, we revisit this paradigm through a careful study on how different samples contribute to the overall performance measured in terms of mAP. Our study suggests that the samples in each mini-batch are neither independent nor equally important, and therefore a better classifier on average does not necessarily mean higher mAP. Motivated by this study, we propose the notion of Prime Samples, those that play a key role in driving the detection performance. We further develop a simple yet effective sampling and learning strategy called PrIme Sample Attention (PISA) that directs the focus of the training process towards such samples. Our experiments demonstrate that it is often more effective to focus on prime samples than hard samples when training a detector. Particularly, On the MSCOCO dataset, PISA outperforms the random sampling baseline and hard mining schemes, e.g., OHEM and Focal Loss, consistently by around 2% on both single-stage and two-stage detectors, even with a strong backbone ResNeXt-101.

References in corpus (8)

• 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
• MMDetection: Open MMLab Detection Toolbox and Benchmark
• Path Aggregation Network for Instance Segmentation
• Hybrid Task Cascade for Instance Segmentation
• Region Proposal by Guided Anchoring
• Deep Regionlets for Object Detection

Cited by in corpus (14)

• Generalized Focal Loss: Learning Qualified and Distributed Bounding Boxes for Dense Object Detection
• Side-Aware Boundary Localization for More Precise Object Detection
• IoU-aware Single-stage Object Detector for Accurate Localization
• IoU-uniform R-CNN: Breaking Through the Limitations of RPN
• Adaptive Instance Distillation for Object Detection in Autonomous Driving
• Learning from Noisy Anchors for One-stage Object Detection
• Pseudo-IoU: Improving Label Assignment in Anchor-Free Object Detection
• Learning a Unified Sample Weighting Network for Object Detection
• Foreground-Background Imbalance Problem in Deep Object Detectors: A Review
• Resisting Crowd Occlusion and Hard Negatives for Pedestrian Detection in the Wild
• The MIS Check-Dam Dataset for Object Detection and Instance Segmentation Tasks
• Prime-Aware Adaptive Distillation
• Feature Super-Resolution Based Facial Expression Recognition for Multi-scale Low-Resolution Faces
• Delving into the Imbalance of Positive Proposals in Two-stage Object Detection