Pedestrian fine-grained attribution detection has drawn significant interest in areas such as smart video surveillance analysis and pedestrian re-identification. Current object detection methods based on deep convolutional neural networks (CNNs) have achieved great progress. However, detections of small parts of pedestrian are still challenging due to their limited resolution and information in images. In this paper, we propose a novel CNN-based framework for improving the accuracy of small objects detection, dubbed MSF-ACNN, using atrous convolutions in cascade along with multi-scale feature fusion: 1) Atrous convolution effectively expands the field-of-view of small regions without increasing the number of parameters and computation. 2) Multi-scale feature fusion obtains more meaningful fine-grained information from both the low-level and high-level feature maps and can handle a variety of image scales. Our results show that MSF-ACNN can obtain better mean average precision (mAP) than the current state-of-the-art methods with faster detection speed, achieving significant improvements on certain small parts of pedestrian such as shoes, bag and hat.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.