Real-time foreground detection approach based on adaptive ensemble learning with arbitrary algorithms for changing environments

Chan, Yi-Tung; Wang, Shuenn-Jyi; Tsai, Chien‐Hsiung

doi:10.1016/j.inffus.2017.05.001

Cited by 17 publications

(7 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Following this idea, different MCSs have been designed in various domains. For example, in computer vision, Chan et al [16] constructed an MCS based on adaptive weighted fusion to combine arbitrary algorithms to detect real-time foreground in changing environments.…”

Section: Challengementioning

confidence: 99%

A DES-based group decision model for group decision making with large-scale alternatives

Liu

Chen

2021

Appl Intell

View full text Add to dashboard Cite

To solve group decision making problems with large-scale alternatives, this paper proposes a dynamic ensemble selection (DES) based group decision model by using historical decision data. The historical decision data of a group of experts are collected from the same multi-criteria decision framework and are mixed to train a set of base classifiers (BCs) to learn group preferences. For each new alternative, the predictions derived from BCs are used to determine its similar historical alternatives from historical data, and the BC with the highest accuracy in predicting the similar historical alternatives is identified as the best individual BC for the new alternative. By iteratively comparing the accuracy of an ensemble of randomly selected BCs and the best individual BC in predicting the similar historical alternatives of the new alternative, a novel DES method is developed to select a competent subset of BCs for the new alternative. The developed DES method effectively avoids the error-independence assumption to a certain extent. Based on the similar historical alternatives determined by the ensemble of selected BCs, a group decision optimization model is developed to learn criterion weights from their assessments on criteria and ensemble predictions derived from the selected BCs. With the learned criterion weights, the understandable group decision result is generated for the new alternative. Case study validates the superiority of the proposed model in diagnosing thyroid nodules using group capabilities. Empirical comparisons on thirty real datasets examine the competence of the proposed DES method against five representative DES methods.

show abstract

Section: Challengementioning

confidence: 99%

A DES-based group decision model for group decision making with large-scale alternatives

Liu

Chen

2021

Appl Intell

View full text Add to dashboard Cite

show abstract

“…It can be clearly seen that the proposed technique outperforms the second best with a huge increment of 9% in Fm Av for 2014 CDnet dataset. Table 5 presents an overall comparison of the proposed technique with recent state-of-the-art techniques, namely, M 4 CD [36], ESILBP [39], RTSS Vibe+PCPNet [35], MELD [55], WeSamBE [42], SuBSENSE [40], PAWCS [41], Ivibe [33], MBS [25], Spectral-360 [56], AMBER [46], AAPSA [47], Vibe [28]. Approaches like IUTIS [57] offering better performance measures by combining different change detection methods have been discarded.…”

Section: Evaluationsmentioning

confidence: 99%

Compute-Extensive Background Subtraction for Efficient Ghost Suppression

2019

View full text Add to dashboard Cite

Efficient background modelling has always been an active area of research due to its immense importance as a preliminary step in various machine-vision applications. Several techniques have been proposed to date that strive to achieve higher accuracy without compromising on computational and hardware demands. One of such techniques, Visual Background Extractor (Vibe), has set benchmarks due to its fewer memory requirements and good results. However, it suffers from high false positives due to its slower, selective and random update policy. This paper proposes a novel sample-consensus pixel-based technique for efficient foreground segmentation complemented with faster ghost suppression. This is achieved by employing segmentation masks exploiting both static and dynamic properties of pixels depicting likeliness towards absorption into the foregrounds. Dynamic characteristics of the proposed approach handle 'object present in the first frame' problem while static characteristics handle improper illumination and shadows in videos in lesser time. It aims not only at suppressing ghosts in the foreground mask but also allows their absorption by updating the background model with such regions. It also proposes a unique spatio-temporal model initialization technique for handling continuous noise. The proposed approach proved to produce outstanding results when compared with 9 traditional and 13 state-of-the-art algorithms. INDEX TERMS Background subtraction, motion detection, Vibe.

show abstract

“…These images contain hundreds of spectral bands and has a high spectral resolution and can thus maintain the subtle differences of ground object features in the spectral dimension. 1 This characteristic makes it quite useful in some applications, such as environmental monitoring, 2,3 precision agriculture, 4,5 and ground objects classification. 6,7 However, high dimensionality of HSIs raises challenges in image processing and use.…”

Section: Introductionmentioning

confidence: 99%

Improved interactive color visualization approach for hyperspectral images

2021

Information Visualization

View full text Add to dashboard Cite

Hyperspectral images (HSIs) have become increasingly prominent as they can maintain the subtle spectral differences of the imaged objects. Designing approaches and tools for analyzing HSIs presents a unique set of challenges due to their high-dimensional characteristics. An improved color visualization approach is proposed in this article to achieve communication between users and HSIs in the field of remote sensing. Under the real-time interactive control and color visualization, this approach can help users intuitively obtain the rich information hidden in original HSIs. Using the dimensionality reduction (DR) method based on band selection, high-dimensional HSIs are reduced to low-dimensional images. Through drop-down boxes, users can freely specify images that participate in the combination of RGB channels of the output image. Users can then interactively and independently set the fusion coefficient of each image within an interface based on concentric circles. At the same time, the output image will be calculated and visualized in real time, and the information it reflects will also be different. In this approach, channel combination and fusion coefficient setting are two independent processes, which allows users to interact more flexibly according to their needs. Furthermore, this approach is also applicable for interactive visualization of other types of multi-layer data.

show abstract

Real-time foreground detection approach based on adaptive ensemble learning with arbitrary algorithms for changing environments

Cited by 17 publications

References 22 publications

A DES-based group decision model for group decision making with large-scale alternatives

A DES-based group decision model for group decision making with large-scale alternatives

Compute-Extensive Background Subtraction for Efficient Ghost Suppression

Improved interactive color visualization approach for hyperspectral images

Contact Info

Product

Resources

About