32,760 Hits in 3.8 sec

Salient Object Ranking with Position-Preserved Attention [article]

Hao Fang, Daoxin Zhang, Yi Zhang, Minghao Chen, Jiawei Li, Yao Hu, Deng Cai, Xiaofei He
2021 arXiv   pre-print
Considering the importance of position in saliency comparison, we preserve absolute coordinates of objects in ROI pooling operation and then fuse positional information with semantic features in the first  ...  We also introduce a Position-Preserved Attention (PPA) module tailored for the SOR branch. It consists of the position embedding stage and feature interaction stage.  ...  The framework performs instance segmentation and salient object ranking simultaneously. We also propose a Position-Preserved Attention (PPA) module tailored for the SOR branch.  ... 
arXiv:2106.05047v2 fatcat:bkmvkkcn4fflhatv665qi5cnqe

A Survey on Visual Saliency Detection and Computational Methods

Ankit Garg, Ashish Negi
2017 International Journal of Engineering and Technology  
In image resizing accurate saliency map helpful to find out accurate salient object that must be preserve during image resizing in this way improved algorithm can be developed with accurate result where  ...  Saliency region detection and object segmentation is also known as salient object detection. To preserve structure of important objects present in the image saliency detection algorithm are used.  ...  Clusters with less positional distance are considered as salient cluster and conversely clusters with high positional distant are considered as less salient. B.  ... 
doi:10.21817/ijet/2017/v9i4/170904406 fatcat:23qqpjjapbde7beaj7fm2myfrm

From rareness to compactness: Contrast-aware image saliency detection

Hsin-Ho Yeh, Chu-Song Chen
2012 2012 19th IEEE International Conference on Image Processing  
Then, the salient regions are detected by aggregating the surrounding regions of the spots, which fulfil the compactness nature of salient objects.  ...  In our approach, multiple-salient-spots are used to find initial salient clues, which appear to be rare and unique parts in an image.  ...  [5] present a global-based approach that preserves a reasonable range of frequency to detect the whole salient object instead of object border.  ... 
doi:10.1109/icip.2012.6467050 dblp:conf/icip/YehC12 fatcat:eqmpka2ugjcj5ikw2gk5ee523a

Infrared and visible image fusion using Latent Low-Rank Representation [article]

Hui Li, Xiao-Jun Wu
2022 arXiv   pre-print
Compared with other fusion methods experimentally, the proposed method has better fusion performance than state-of-the-art fusion methods in both subjective and objective evaluation.  ...  Then, the low-rank parts are fused by weighted-average strategy to preserve more contour information.  ...  We use both subjective and objective methods to evaluate the proposed method, the experimental results show that the proposed method exhibits better performance than other compared methods.  ... 
arXiv:1804.08992v5 fatcat:xo766fmw2jf63eu2blhvjmo56u

Instance-Level Relative Saliency Ranking with Graph Reasoning [article]

Nian Liu, Long Li, Wangbo Zhao, Junwei Han, Ling Shao
2021 arXiv   pre-print
Conventional salient object detection models cannot differentiate the importance of different salient objects.  ...  However, one of these models cannot differentiate object instances and the other focuses more on sequential attention shift order inference.  ...  The most salient objects are well preserved with less deformation, while less salient objects can be deformed first or removed under very large-scale reduction.  ... 
arXiv:2107.03824v1 fatcat:2hhm6p3jv5h2fhtiefrpm55aee

Global and Local Sensitivity Guided Key Salient Object Re-augmentation for Video Saliency Detection [article]

Ziqi Zhou, Zheng Wang, Huchuan Lu, Song Wang, Meijun Sun
2018 arXiv   pre-print
KSORA includes two sub-modules (WFE and KOS): WFE processes local salient feature selection using bottom-up strategy, while KOS ranks each object in global fashion by top-down statistical knowledge, and  ...  In this paper, based on the fact that salient areas in videos are relatively small and concentrated, we propose a key salient object re-augmentation method (KSORA) using top-down semantic knowledge and  ...  positioning of the key salient object, which prove the validity of our proposed model.  ... 
arXiv:1811.07480v1 fatcat:xeypmi5u7zhzpay7s2rplnxcry

Deep Progressive Hashing for Image Retrieval

Jiale Bai, Bingbing Ni, Minsi Wang, Yang Shen, Hanjiang Lai, Chongyang Zhang, Lin Mei, Chuanping Hu, Chen Yao
2017 Proceedings of the 2017 ACM on Multimedia Conference - MM '17  
The proposed deep hashing network is trained via minimizing a triplet ranking loss, which is end-to-end trainable.  ...  Inspired by human's nonsalient-to-salient perception path, the proposed hashing scheme generates a series of binary codes based on progressively expanded salient regions.  ...  Similarity Preserving Objective: Among the current learning-based hashing methods, supervised hashing preserve the pair-wised similarities or triple-wised rankings, devised by the supervised information  ... 
doi:10.1145/3123266.3123280 dblp:conf/mm/BaiNWSLZMHY17 fatcat:zmcavxe6bzcu5dtr4kkyqetnxu

Horizontal-to-Vertical Video Conversion [article]

Tun Zhu, Daoxin Zhang, Yao Hu, Tianran Wang, Xiaolong Jiang, Jianke Zhu, Jiawei Li
2021 arXiv   pre-print
To achieve so, we propose a Rank-SS module that detects human objects, then selects the subject-to-preserve via exploiting location, appearance, and salient cues.  ...  Concretely, H2V framework integrates video shot boundary detection, subject selection and multi-object tracking to facilitate the subject-preserving conversion, wherein the key is subject selection.  ...  TABLE VII : VII The ANOVA experiment of Rank Sub-Select module with Salient Object Detection and Fixation Prediction methods.  ... 
arXiv:2101.04051v2 fatcat:x65xhdepvzb4dhe2yxh27warhq

IDA: Improved Data Augmentation Applied to Salient Object Detection [article]

Daniel V. Ruiz and Bruno A. Krinski and Eduardo Todt
2020 arXiv   pre-print
Our proposed technique enables more precise control of the object's position and size while preserving background information.  ...  Our method combines image inpainting, affine transformations, and the linear combination of different generated background images with salient objects extracted from labeled data.  ...  We gratefully acknowledge the founders of the publicly available datasets and the support of NVIDIA Corporation with the donation of the GPUs used for this research.  ... 
arXiv:2009.08845v1 fatcat:imodzyz7gjda7hypusdu2ibg3m

A New Similarity Measure with Deformation Detection of Visual Salient Regions for Image Retargeting

Canlin Li, Fubao Zhu, Rijian Su
2014 International Journal of Multimedia and Ubiquitous Engineering  
Content similarity for image retargeting mainly involves the number and layout of the salient contents in the image, and visual effect similarity mainly focuses on shape preservation of visual salient  ...  Experimental results show that the objective quality values closely match the subjective scores evaluated by users, indicating that our proposed objective metrics are congruent with human perception mechanism  ...  Experimental results showed that the achieved similarity ranks for image retargeting closely match the subjective human choices, indicating that our proposed objective measure is almost congruent with  ... 
doi:10.14257/ijmue.2014.9.7.01 fatcat:ddojtc3np5b4lifte7lgxnjz6u

Saliency detection: A self-ordinal resemblance approach

Wonjun Kim, Chanho Jung, Changick Kim
2010 2010 IEEE International Conference on Multimedia and Expo  
In saliency detection, regions attracting visual attention need to be highlighted while effectively suppressing non-salient regions for the semantic scene understanding.  ...  To justify robustness of our approach, the proposed method is compared with the state of the art methods on various images. 1  ...  In contrast to that, the proposed saliency map suppresses effectively high energy pixels generated in non-salient regions and thus the shape of important objects is preserved more efficiently.  ... 
doi:10.1109/icme.2010.5583287 dblp:conf/icmcs/KimJK10 fatcat:ud2otm2q6nfcvk6wwrden44itq

An Objective Quality of Experience (QoE) Assessment Index for Retargeted Images

Jiangyang Zhang, C.-C. Jay Kuo
2014 Proceedings of the ACM International Conference on Multimedia - MM '14  
Experimental results demonstrate that the proposed GLS quality index has stronger correlation with human QoE than other existing objective metrics in retargeted image quality assessment.  ...  They are global structural distortion (G), local region distortion (L) and loss of salient information (S). Different features are selected to quantify their respective distortion degrees.  ...  As shown in the rank order table of Fig. 11 , the objective rank computed with the proposed GLS index correlates well with the subjective rank in [17] .  ... 
doi:10.1145/2647868.2654922 dblp:conf/mm/ZhangK14 fatcat:ngh7egfbubblfnzapbynt2spm4

Salient Object Detection: A Benchmark

Ali Borji, Ming-Ming Cheng, Huaizu Jiang, Jia Li
2015 IEEE Transactions on Image Processing  
the purpose of benchmarking salient object detection and segmentation methods.  ...  We extensively compare, qualitatively and quantitatively, 40 state-of-the-art models (28 salient object detection, 10 fixation prediction, 1 objectness, and 1 baseline) over 6 challenging datasets for  ...  Firstly, it detects the most salient and attention-grabbing object in a scene, and then it segments the whole extent of that object.  ... 
doi:10.1109/tip.2015.2487833 pmid:26452281 fatcat:bkoyk3izxbczpigbpfwqfypjre

Salient Object Detection: A Benchmark [chapter]

Ali Borji, Dicky N. Sihite, Laurent Itti
2012 Lecture Notes in Computer Science  
Saliency models that intend to predict eye fixations perform lower on segmentation datasets compared to salient object detection algorithms.  ...  Several salient object detection approaches have been published which have been assessed using different evaluation scores and datasets resulting in discrepancy in model comparison.  ...  Firstly, it detects the most salient and attention-grabbing object in a scene, and then it segments the whole extent of that object.  ... 
doi:10.1007/978-3-642-33709-3_30 fatcat:vswe3vvqi5dv7dsro3we6xcu3q

Invariance Analysis of Saliency Models versus Human Gaze During Scene Free Viewing [article]

Zhaohui Che, Ali Borji, Guangtao Zhai, Xiongkuo Min
2018 arXiv   pre-print
Despite few efforts, influences of ubiquitous distortions on visual attention and saliency models have not been systematically investigated.  ...  and saliency models, we find that: a) observers look at different locations over distorted versus original images, and b) performances of saliency models are drastically hindered over distorted images, with  ...  We will share our collected data and code with the community to promote research in improving the robustness of deep models over different distortions and to close the gap between saliency models and the  ... 
arXiv:1810.04456v1 fatcat:gurkaxdf7vbrtmj5k3m5z2evzi
« Previous Showing results 1 — 15 out of 32,760 results