Filters








4,905 Hits in 4.8 sec

Visual Saliency Detection Based on Multiscale Deep CNN Features

Guanbin Li, Yizhou Yu
2016 IEEE Transactions on Image Processing  
To generate a more robust feature, we integrate handcrafted low-level features with our deep contrast feature.  ...  The penultimate layer of our neural network has been confirmed to be a discriminative high-level feature vector for saliency detection, which we call deep contrast feature.  ...  ACKNOWLEDGMENT The authors would like to thank Sai Bi, Wei Zhang, and Feida Zhu for their help during the construction of our dataset.  ... 
doi:10.1109/tip.2016.2602079 pmid:28113629 fatcat:enyhpxlk5ffarnnf57zshjhyya

RGBD Salient Object Detection via Deep Fusion

Liangqiong Qu, Shengfeng He, Jiawei Zhang, Jiandong Tian, Yandong Tang, Qingxiong Yang
2017 IEEE Transactions on Image Processing  
In this paper, we design a new convolutional neural network (CNN) to fuse different low level saliency cues into hierarchical features for automatically detecting salient objects in RGBD images.  ...  Numerous efforts have been made to design different low level saliency cues for the RGBD saliency detection, such as color or depth contrast features, background and color compactness priors.  ...  In this paper, we design a new convolutional neural network (CNN) to fuse different low level saliency cues into hierarchical features for automatically detecting salient objects in RGBD images.  ... 
doi:10.1109/tip.2017.2682981 pmid:28320666 fatcat:m5jdky3zzrfzbnj3mvjaybf62q

HARF: Hierarchy-Associated Rich Features for Salient Object Detection

Wenbin Zou, Nikos Komodakis
2015 2015 IEEE International Conference on Computer Vision (ICCV)  
To address such an issue, this paper proposes a novel hierarchy-associated feature construction framework for salient object detection, which is based on integrating elementary features from multi-level  ...  from background, largely due to the lack of sufficiently robust features for saliency prediction.  ...  Thus most of the previous saliency models define regional saliency based on the regional contrasts of traditional low-level features, such as color and texture.  ... 
doi:10.1109/iccv.2015.54 dblp:conf/iccv/ZouK15 fatcat:zfs7la6xdzgbpn5swiqs3ms4ve

LCNN: Low-level Feature Embedded CNN for Salient Object Detection [article]

Hongyang Li, Huchuan Lu, Zhe Lin, Xiaohui Shen, Brian Price
2015 arXiv   pre-print
In this paper, we propose a novel deep neural network framework embedded with low-level features (LCNN) for salient object detection in complex images.  ...  Several low-level features are extracted, which can effectively capture contrast and spatial information in the salient regions, and incorporated to compensate with the learned high-level features at the  ...  LCNN for saliency detection We concatenate the low-level feature vector proposed above with the high-level feature vector generated from layer 7 and use them as input of the SVM detector (see Figure 2  ... 
arXiv:1508.03928v1 fatcat:wbmvm5eqije4tm4h5s2a2usjhm

Video Smoke Detection Based on Deep Saliency Network [article]

Gao Xu, Yongming Zhang, Qixing Zhang, Gaohua Lin, Zhong Wang, Yang Jia, Jinjun Wang
2019 arXiv   pre-print
The deep feature map is combined with the saliency map to predict the existence of smoke in an image.  ...  An end-to-end framework for salient smoke detection and existence prediction of smoke is proposed for application in video smoke detection.  ...  Li [19] integrates handcrafted low-level features with deep contrast features for a more robust feature.  ... 
arXiv:1809.02802v2 fatcat:lxewqciajrai7b3zeybgsykak4

Visualization of Salient Object with Saliency Maps using Residual Neural Networks

Rubina Rashid, Saqib Ubaid, Muhammad Idrees, Rida Rafi, Imran Sarwar Bajwa
2021 IEEE Access  
In computer vision, visual saliency detection is one of the main challenges and CNNs are most powerful techniques that are used widely for different layers integration to make saliency maps [38] .  ...  Commonly, visual saliency models use multi-scales configuration for improving accuracy which integrates the information at low and high image scales [11] .  ... 
doi:10.1109/access.2021.3100155 fatcat:aqzrky7vhfbfbpwazj4fva2w2q

Deep Saliency with Encoded Low level Distance Map and High Level Features [article]

Gayoung Lee, Yu-Wing Tai, Junmo Kim
2016 arXiv   pre-print
Our method utilizes both high level and low level features for saliency detection under a unified deep learning framework.  ...  These advances have demonstrated superior results over previous works that utilize hand-crafted low level features for saliency detection.  ...  Conclusion In this paper, we have introduced a new method to integrate the low-level and the high-level features for saliency detection.  ... 
arXiv:1604.05495v1 fatcat:n25doge3onfcdk543w7zhf7l4y

Deep Saliency with Encoded Low Level Distance Map and High Level Features

Gayoung Lee, Yu-Wing Tai, Junmo Kim
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
Our method utilizes both high level and low level features for saliency detection under a unified deep learning framework.  ...  These advances have demonstrated superior results over previous works that utilize hand-crafted low level features for saliency detection.  ...  Conclusion In this paper, we have introduced a new method to integrate the low-level and the high-level features for saliency detection.  ... 
doi:10.1109/cvpr.2016.78 dblp:conf/cvpr/LeeTK16 fatcat:p7gtsecv7ve3tovtxh2swxliga

Multichannel Fusion Based on modified CNN for Image Emotion Recognition

Juntao Zhao Juntao Zhao
2022 Diànnǎo xuékān  
Therefore, we propose a multichannel fusion method based on modified CNN for image emotion recognition. Firstly, saliency target and face target region are detected in the whole image.  ...  Then feature pyramid is used to improve CNN to recognize saliency target emotion. Weighted loss CNN emotion recognition is constructed on multi-layer supervision module.  ...  Generally, CNN is used to extract high-level semantic information of foreground image and low-level basic features of background image respectively.  ... 
doi:10.53106/199115992022023301002 fatcat:fpwl72j56nas3gt3b3estdv4r4

Saliency Detection by Forward and Backward Cues in Deep-CNNs [article]

Nevrez Imamoglu, Chi Zhang, Wataru Shimoda, Yuming Fang, Boxin Shi
2017 arXiv   pre-print
In this paper, we propose a top-down saliency model using CNN, a weakly supervised CNN model trained for 1000 object labelling task from RGB images.  ...  As prior knowledge of objects or object features helps us make relations for similar objects on attentional tasks, pre-trained deep convolutional neural networks (CNNs) can be used to detect salient objects  ...  So, in addition to the CNN model learning, a secondary supervision on CNN features to the pixel level ground truth saliency maps can be done.  ... 
arXiv:1703.00152v2 fatcat:6kqspoaz4bdtzoefcx3zlbwz2y

Visual saliency detection for RGB-D images under a Bayesian framework

Songtao Wang, Zhen Zhou, Wei Jin, Hanbing Qu
2018 IPSJ Transactions on Computer Vision and Applications  
In this paper, we propose a saliency detection model for RGB-D images based on the deep features of RGB images and depth images within a Bayesian framework.  ...  By analysing 3D saliency in the case of RGB images and depth images, the class-conditional mutual information is computed for measuring the dependence of deep features extracted using a convolutional neural  ...  From the above description, the key to 3D saliency detection models is determining how to integrate the depth cues with traditional 2D low-level features.  ... 
doi:10.1186/s41074-017-0037-0 fatcat:jklxzq46vrbh3mk6n7ocbqhoiu

Accurate salient object detection via dense recurrent connections and residual-based hierarchical feature integration

Yanpeng Cao, Guizhong Fu, Jiangxin Yang, Yanlong Cao, Michael Ying Yang
2019 Signal processing. Image communication  
In this paper, we propose a novel CNN-based saliency detection method through dense recurrent connections and residual-based hierarchical feature integration.  ...  Then we present a residual-based architecture with short connections for deep supervision which hierarchically combines both coarse-level and fine-level feature representations.  ...  It is reasonable to utilize both high-level and low-level features extracted in layers with different depths for robust and accurate saliency detection.  ... 
doi:10.1016/j.image.2019.06.004 fatcat:k5dxyhyu7vdzrccq3o7zpdkwlq

Multi-Scale Global Contrast CNN for Salient Object Detection

Weijia Feng, Xiaohui Li, Guangshuai Gao, Xingyue Chen, Qingjie Liu
2020 Sensors  
In contrast to many previous CNN based saliency methods that apply super-pixel segmentation to obtain homogeneous regions and then extract their CNN features before producing saliency maps region-wise,  ...  In this paper, we design an end-to-end multi-scale global contrast convolutional neural network (CNN) that explicitly learns hierarchical contrast information among global and local features of an image  ...  Acknowledgments: We are grateful to the reviewers for their valuable comments and suggestions that help us to improve this work. We also thank authors who kindly provide their codes for comparison.  ... 
doi:10.3390/s20092656 pmid:32384766 pmcid:PMC7248752 fatcat:6xuv47dm3jfptl4ckvermbtjnu

An integration of bottom-up and top-down salient cues on RGB-D data: saliency from objectness versus non-objectness

Nevrez Imamoglu, Wataru Shimoda, Chi Zhang, Yuming Fang, Asako Kanezaki, Keiji Yanai, Yoshifumi Nishida
2017 Signal, Image and Video Processing  
RGB-D saliency integration framework yields promising results compared with the several state-of-the-art-models.  ...  We demonstrate that combining salient features from color and dept through bottom-up and top-down methods gives significant improvement on the salient object detection with space based and object based  ...  Boxin Shi at AIST (Tokyo, Japan) for discussions on 3D data processing.  ... 
doi:10.1007/s11760-017-1159-7 fatcat:b5nr2bj7bzbbzb7nmkdrmnseiq

Multi-Scale Cascade Network for Salient Object Detection

Xin Li, Fan Yang, Hong Cheng, Junyu Chen, Yuxiao Guo, Leiting Chen
2017 Proceedings of the 2017 ACM on Multimedia Conference - MM '17  
Our network consists of several stages (sub-networks) for handling saliency detection across different scales.  ...  Compared with existing CNN-based saliency models, the MSC-Net can naturally enable the learning process in the finer cascade stages to encode more global contextual information while progressively incorporating  ...  This combination enriches the underlying representations with more information (both low-level and high-level features), which is helpful for salient object detection.  ... 
doi:10.1145/3123266.3123290 dblp:conf/mm/LiYCCGC17 fatcat:ssbtm2omszeythyqkjed5p6ajq
« Previous Showing results 1 — 15 out of 4,905 results