期刊
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
卷 31, 期 2, 页码 715-727出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TCSVT.2020.2987465
关键词
Convolutional neural networks (CNNs); deep learning; object detection; object recognition
资金
- National Natural Science Foundation of China [61525102, 61831005, 61971095, 61871078]
This paper introduces PDC-Net, a multi-path detection calibration network, to address the data distribution discrepancy between object proposals and refined bounding-boxes. Built on Faster R-CNN, PDC-Net utilizes a multi-path detection head to calibrate detection results and improve accuracy.
Object proposals are used in two-stage detectors, such as R-CNN, to generate detection results, including category predictions and refined bounding-boxes. As a result, classification scores are assigned to refined bounding-boxes rather than object proposals. However, this procedure ignores the discrepancy of data distribution between object proposals and refined bounding-boxes. We consider this discrepancy could limit the detection accuracy. Specifically, the foreground/background imbalance on object proposals and inaccurate information from low-IoU proposals could hinder the category prediction. In this paper, we propose a detector called the Multi-Path Detection Calibration Network (PDC-Net) to address this problem. The key idea behind PDC-Net is calibrating detection results from R-CNN by considering the statistical discrepancy between object proposals and refined bounding-boxes. PDC-Net is built on Faster R-CNN. The core component in PDC-Net is the multi-path detection head, in which the base detector (from Faster R-CNN) generates detection results from object proposals and multiple calibration detectors fix incorrect outputs from the base detector using refined bounding-boxes. Experiments reveal that PDC-Net can boost detection results. Our method could reach 83.1% and 43.3% mAP respectively on PASCAL VOC and MSCOCO benchmarks, which is comparable to several state-of-the-art methods.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据