Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Kim, Jae-Yeul | - |
dc.contributor.author | Ha, Jong-Eun | - |
dc.date.accessioned | 2022-11-07T07:30:32Z | - |
dc.date.available | 2022-11-07T07:30:32Z | - |
dc.date.created | 2022-10-28 | - |
dc.date.issued | 2022-10 | - |
dc.identifier.issn | 2169-3536 | - |
dc.identifier.uri | http://hdl.handle.net/20.500.11750/17049 | - |
dc.description.abstract | In visual surveillance, deep learning-based foreground object detection algorithms are superior to classical background subtraction (BGS)-based algorithms. However, deep learning-based methods are limited because detection performance deteriorates in a new environment different from the training environment. This limitation can be solved by retraining the model using additional ground-truth labels in the new environment. However, generating ground-truth labels for visual surveillance is time-consuming and expensive. This paper proposes a method that does not require foreground labels when adapting to a new environment. To this end, we propose an integrated network that produces two kinds of outputs a background model image and a foreground object map. We can adapt to the new environment by retraining using a background model image. The proposed method consists of one encoder and two decoders for detecting foreground objects and a background model image. It is designed to enable real-time processing with desktop GPUs. The proposed method shows 14.46% improved FM in a new environment different from training and 11.49% higher FM than the latest BGS algorithm. | - |
dc.language | English | - |
dc.publisher | Institute of Electrical and Electronics Engineers Inc. | - |
dc.title | Weakly Supervised Foreground Object Detection Network Using Background Model Image | - |
dc.type | Article | - |
dc.identifier.doi | 10.1109/ACCESS.2022.3211987 | - |
dc.identifier.scopusid | 2-s2.0-85140195770 | - |
dc.identifier.bibliographicCitation | IEEE Access, v.10, pp.105726 - 105733 | - |
dc.description.isOpenAccess | TRUE | - |
dc.subject.keywordAuthor | Supervised learning | - |
dc.subject.keywordAuthor | Visualization | - |
dc.subject.keywordAuthor | Surveillance | - |
dc.subject.keywordAuthor | Feature extraction | - |
dc.subject.keywordAuthor | Object detection | - |
dc.subject.keywordAuthor | Decoding | - |
dc.subject.keywordAuthor | Data models | - |
dc.subject.keywordAuthor | Deep learning | - |
dc.subject.keywordAuthor | Visual surveillance | - |
dc.subject.keywordAuthor | weakly supervised | - |
dc.subject.keywordAuthor | deep learning | - |
dc.subject.keywordAuthor | foreground object detection | - |
dc.citation.endPage | 105733 | - |
dc.citation.startPage | 105726 | - |
dc.citation.title | IEEE Access | - |
dc.citation.volume | 10 | - |
There are no files associated with this item.