2021
DOI: 10.1109/tmm.2020.3019688
|View full text |Cite
|
Sign up to set email alerts
|

Saliency Detection Using Deep Features and Affinity-Based Robust Background Subtraction

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
8
0

Year Published

2021
2021
2023
2023

Publication Types

Select...
7
1

Relationship

0
8

Authors

Journals

citations
Cited by 24 publications
(8 citation statements)
references
References 52 publications
0
8
0
Order By: Relevance
“…XAI systems have been developed to meet this challenge, primarily motivated by image classification concerns like the erroneous classifications with the ruler in the image problem [79]. One example of an XAI method that opens the AI black box for interpretability is a saliency map [81]. A saliency map reveals information on the degree that each feature in the image explain and contribute to predictions [82].…”
Section: Application Of Explainable Models Goals Of Explainable Aimentioning
confidence: 99%
“…XAI systems have been developed to meet this challenge, primarily motivated by image classification concerns like the erroneous classifications with the ruler in the image problem [79]. One example of an XAI method that opens the AI black box for interpretability is a saliency map [81]. A saliency map reveals information on the degree that each feature in the image explain and contribute to predictions [82].…”
Section: Application Of Explainable Models Goals Of Explainable Aimentioning
confidence: 99%
“…Researchers have found that fusing multiple sensor data with RGB data can potentially improve object classification performance [51][52][53]. Commonly used fusion approaches can be categorized into four types: input, early, late, and multiscale fusion.…”
Section: Image Data Fusion and Related Workmentioning
confidence: 99%
“…Nevertheless, their fusion methods provide alternative solutions to segmentation problems. For example, Nawaz et al [51] built attention map networks (AMNs) in addition to traditional feature extraction methods to subtract background information from RGB images. In their method, the attention map is a mechanism that extracts different weights from original RGB images.…”
Section: Rgb-depth Fusionmentioning
confidence: 99%
“…Foreground–background separation 7 9 aims to separate a video into a static background sequence (low-rank) and a moving foreground sequence (sparse). Given a video with n single-frame images of size w×h, we vectorize each frame as a vector ddouble-struckRm (m=w×h).…”
Section: Introductionmentioning
confidence: 99%
“…Low-rank and sparse decomposition (LRSD), 1,2 which refers to the problem of principal component analysis in the presence of outliers, recovers the hidden low-rank matrix L and the sparse matrix S from its observation matrix D and is an important first step in simplifying many video analytics tasks. [3][4][5][6] Foreground-background separation [7][8][9] aims to separate a video into a static background sequence (low-rank) and a moving foreground sequence (sparse). Given a video with n singleframe images of size w × h, we vectorize each frame as a vector d ∈ R m (m ¼ w × h).…”
Section: Introductionmentioning
confidence: 99%