Saliency-based selection of visual content for deep convolutional neural networks

Obeso, Abraham Montoya; Benois-Pineau, Jenny; Vázquez, Mireya Saraí García; Acosta, Alejandro

doi:10.1007/s11042-018-6515-2

Cited by 15 publications

(10 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Zhao et al [68] offered a survey on deep learning methods for object detection using visual saliency approaches. Obeso et al [69] used a saliencydriven approach to predict visual attention in images and use it to train a Deep Convolutional Neural Network. Huang et al [70] proposed a method based on the adaptation of deep neural networks.…”

Section: ) Deep Neural Network Based Saliency Techniquesmentioning

confidence: 99%

A Multi-Scale Colour and Keypoint Density-Based Approach for Visual Saliency Detection

et al. 2020

View full text Add to dashboard Cite

In the first seconds of observation of an image, several visual attention processes are involved in the identification of the visual targets that pop-out from the scene to our eyes. Saliency is the quality that makes certain regions of an image stand out from the visual field and grab our attention. Saliency detection models, inspired by visual cortex mechanisms, employ both colour and luminance features. Furthermore, both locations of pixels and presence of objects influence the Visual Attention processes. In this paper, we propose a new saliency method based on the combination of the distribution of interest points in the image with multiscale analysis, a centre bias module and a machine learning approach. We use perceptually uniform colour spaces to study how colour impacts on the extraction of saliency. To investigate eye-movements and assess the performances of saliency methods over object-based images, we conduct experimental sessions on our dataset ETTO (Eye Tracking Through Objects). Experiments show our approach to be accurate in the detection of saliency concerning state-of-the-art methods and accessible eye-movement datasets. The performances over object-based images are excellent and consistent on generic pictures. Besides, our work reveals interesting findings on some relationships between saliency and perceptually uniform colour spaces.

show abstract

Section: ) Deep Neural Network Based Saliency Techniquesmentioning

confidence: 99%

A Multi-Scale Colour and Keypoint Density-Based Approach for Visual Saliency Detection

et al. 2020

View full text Add to dashboard Cite

show abstract

“…The loss functional are indicates as. τ T otal = ατ spatial + βτ temporal (7) Where α and β is a learnable balance factor. Summary ,we can better perceive changes in the content of urban and rural intentions in this way, achieve as much as possible the automated processing of the content of urban and rural intentions, and improve the ability and efficiency of emergency response after disasters.…”

Section: Optimizationmentioning

confidence: 99%

“…[5] [6]. Obeso, A Montoya and Benois-Pineau et.al [7] to addresses the classification problem of Mexican cultural heritage adopt a deep convolution netural networks methods to training and predict visual attention in natural images.Morbidoni, Christian and Pierdicca et.al [8] proposed a novelty methods of learning from synthetic point cloud data for historical buildings semantic segmentation, where the mainlyto provide a first assessment of the use of synthetic data to drive convolution netural networks based semantic segmentation in the context of historical buildings.…”

Section: Introductionmentioning

confidence: 99%

Exploring Spatiotemporal Change of City and Village from Remote Sensing Using Multi-Branch Networks

Zhao

Tian

2021

Preprint

View full text Add to dashboard Cite

With the rapid development of social economy, the urban and rural environment, form and infrastructure have also under- gone earth-shaking changes. As a gathering place for human activities, urban and rural areas play a vital role in the interaction between humans and society. If traditional machine learning methods are used to perceive changes in the intentional connotation of urban and rural areas, it is easy to ignore the detailed information of the intentional target. At the same time, the perception accuracy needs to be improved. Therefore, the deep neural network in this paper proposes a way to perceive the temporal and spatial changes of urban and rural intentional connotations from the perspective of remote sensing. The framework first uses the multi-branch DenseNet to capture the multi-scale spatiotemporal information of the intended target, and embeds it with high-level semantics and low-level physical appearance information. Secondly, a multi-branch cross-channel attention module is designed to perform reﬁned and integrated processing of multi-level spatiotemporal information to realize the aggregation of multi-branch information. Finally, it was tested and verified on two different time data sets, and good performance was achieved.

show abstract

“…where denotes velocity, the momentum coefficient ( = 0.9), the learning rate, the iteration and the filters coefficients. The exponential learning rate decay policy proved to be efficient in works with also not a large number of training data available [33]. We thus use it : = 0 .…”

Section: Architecture Choice and Tunning Of Cnn For Risk Situations Rmentioning

confidence: 99%

Detection of Semantic Risk Situations in Lifelog Data for Improving Life of Frail People

Yebda

Benois‐Pineau

Pech

et al. 2020

Proceedings of the 2020 International Conference on Multimedia Retrieval

View full text Add to dashboard Cite

The automatic recognition of risk situations for frail people is an urgent research topic for the interdisciplinary artificial intelligence and multimedia community. Risky situations can be recognized from lifelog data recorded with wearable devices. In this paper, we present a new approach for the detection of semantic risk situations for frail people in lifelog data. Concept matching between general lifelog and risk taxonomies was realized and tuned AlexNet was deployed for detection of two semantic risks situations such as risk of domestic accident and risk of fraud with promising results. CCS CONCEPTS • Applied computing → Health care information systems; • Computing methodologies → Neural networks.

show abstract

Saliency-based selection of visual content for deep convolutional neural networks

Cited by 15 publications

References 39 publications

A Multi-Scale Colour and Keypoint Density-Based Approach for Visual Saliency Detection

A Multi-Scale Colour and Keypoint Density-Based Approach for Visual Saliency Detection

Exploring Spatiotemporal Change of City and Village from Remote Sensing Using Multi-Branch Networks

Detection of Semantic Risk Situations in Lifelog Data for Improving Life of Frail People

Contact Info

Product

Resources

About