期刊
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
卷 32, 期 5, 页码 2075-2089出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TNNLS.2020.2996406
关键词
Benchmark; RGB-D; saliency; salient object detection (SOD); Salient Person (SIP) data set
类别
资金
- Major Project for New Generation of AI [2018AAA0100400]
- NSFC [61922046]
- Tianjin Natural Science Foundation [17JCJQJC43700]
This article makes contributions to RGB-D SOD by collecting a new SIP dataset, conducting a large-scale benchmark comparing contemporary methods, and proposing the D(3)Net model. D(3)Net outperforms prior contenders and can efficiently extract salient object masks for real scenes.
The use of RGB-D information for salient object detection (SOD) has been extensively explored in recent years. However, relatively few efforts have been put toward modeling SOD in real-world human activity scenes with RGB-D. In this article, we fill the gap by making the following contributions to RGB-D SOD: 1) we carefully collect a new Salient Person (SIP) data set that consists of similar to 1 K high-resolution images that cover diverse real-world scenes from various viewpoints, poses, occlusions, illuminations, and backgrounds; 2) we conduct a large-scale (and, so far, the most comprehensive) benchmark comparing contemporary methods, which has long been missing in the field and can serve as a baseline for future research, and we systematically summarize 32 popular models and evaluate 18 parts of 32 models on seven data sets containing a total of about 97k images; and 3) we propose a simple general architecture, called deep depth-depurator network (D(3)Net). It consists of a depth depurator unit (DDU) and a three-stream feature learning module (FLM), which performs low-quality depth map filtering and cross-modal feature learning, respectively. These components form a nested structure and are elaborately designed to be learned jointly. D(3)Net exceeds the performance of any prior contenders across all five metrics under consideration, thus serving as a strong model to advance research in this field. We also demonstrate that D(3)Net can be used to efficiently extract salient object masks from real scenes, enabling effective background-changing application with a speed of 65 frames/s on a single GPU. All the saliency maps, our new SIP data set, the D(3)Net model, and the evaluation tools are publicly available at https://github.com/DengPingFan/D3NetBenchmark.
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