271 Hits in 5.0 sec

Thumbnail Selection with Convolutional Neural Network Based on Emotion Detection

Mahmut ÇAKAR, Kazım YILDIZ, Önder DEMİR
2021 International Journal of Advances in Engineering and Pure Sciences  
The thumbnails with the most repeating faces were selected by developing a face recognition model at each step. The experimental results showed that the emotion model was successful.  ...  Also, deep learning is used to label images with objects and emotions based on the identity of the face.  ...  It trains the weight matrices in the fully connected layers and parameters of the network by spreading the loss derivative backward according to the parameters in the network using filters in the convolutional  ... 
doi:10.7240/jeps.900561 fatcat:tcwe5kbywvc5bj6vbm5l4kibqm

Sentence Specified Dynamic Video Thumbnail Generation

Yitian Yuan, Lin Ma, Wenwu Zhu
2019 Proceedings of the 27th ACM International Conference on Multimedia - MM '19  
Specifically, GTP leverages a sentence specified video graph convolutional network to model both the sentence-video semantic interaction and the internal video relationships incorporated with the sentence  ...  In this paper, we define a distinctively new task, namely sentence specified dynamic video thumbnail generation, where the generated thumbnails not only provide a concise preview of the original video  ...  In face of this data deluge, video thumbnail [23, 29] , as a commonly used technology to provide viewers a condensed and straightforward preview about the video contents, is becoming increasingly crucial  ... 
doi:10.1145/3343031.3350985 dblp:conf/mm/YuanM019 fatcat:62i2nctjmjc7ja7metv4pa7k2u

Client-Driven Personalized Trailer Framework Using Thumbnail Containers

Ghulam Mujtaba, Eun-Seok Ryu.
2020 IEEE Access  
This can be achieved by analyzing personal user data, and can lead to two major issues: privacy violation and enormous demand in computational resources.  ...  The system is designed to support a wide range of client hardware with different computational capabilities and has the flexibility to adapt to network conditions.  ...  FIGURE 5 . 5 Convolutional neural network Inception-v3 architecture.  ... 
doi:10.1109/access.2020.2982992 fatcat:nljypbutxrfv5naitdq6o2lddq

DEEP-AD: A Multimodal Temporal Video Segmentation Framework for Online Video Advertising

Ruxandra Tapu, Bogdan Mocanu, Titus Zaharia
2020 IEEE Access  
INDEX TERMS Multimodal temporal video segmentation, thumbnail extraction from video scenes, commercial advertisement insertion based on semantic criterions, deep convolutional neural networks. 99584 VOLUME  ...  The proposed algorithm exploits various deep convolutional neural networks, involved at several stages. The video stream is first divided into shots based on a graph partition method.  ...  In order to determine the scene thumbnail, we apply as input to the network all the scene key-frames in order to rank the images according to their visual representativeness.  ... 
doi:10.1109/access.2020.2997949 fatcat:odg577q6nbcqffnt3ufjw5dol4

Learning Visual Importance for Graphic Designs and Data Visualizations

Zoya Bylinskii, Nam Wook Kim, Peter O'Donovan, Sami Alsheikh, Spandan Madan, Hanspeter Pfister, Fredo Durand, Bryan Russell, Aaron Hertzmann
2017 Proceedings of the 30th Annual ACM Symposium on User Interface Software and Technology - UIST '17  
We present automated models that predict the relative importance of different elements in data visualizations and graphic designs.  ...  Our models are neural networks trained on human clicks and importance annotations on hundreds of designs.  ...  Since we have limited training data we initialized the network parameters with the pre-trained FCN32s model for semantic segmentation in natural images [32] , and fine-tuned it for our task.  ... 
doi:10.1145/3126594.3126653 dblp:conf/uist/BylinskiiKOAMPD17 fatcat:zxxaqb23njfnlfmr2v2naoixgi

An Effective Cloud Detection Method for Gaofen-5 Images via Deep Learning

Junchuan Yu, Yichuan Li, Xiangxiang Zheng, Yufeng Zhong, Peng He
2020 Remote Sensing  
Inspired by the outstanding learning capability of convolutional neural networks (CNNs) for feature extraction, we propose a new dual-branch CNN architecture for cloud segmentation for GF-5 preview RGB  ...  Deep learning technology, however, is able to perform cloud detection efficiently for massive repositories of satellite data and can even dramatically speed up processing by utilizing thumbnails.  ...  Inspired by the excellent CNN architectures of bilateral segmentation network (BiSeNet) [54] , pyramid scene parsing network (PSPNet) [55] , and squeeze-and-excitation network (SENet) [56] , in this  ... 
doi:10.3390/rs12132106 fatcat:zsgxihvylbhu5l3pyxn2bod33i

Application of Artificial Intelligence in detection of diseases in plants: A Survey

Gyan Singh Sujawat
2021 Turkish Journal of Computer and Mathematics Education  
of diseases in crops  ...  Artificial intelligence is having its vast applications in various sectors.  ...  Convolutional neural networks : Convolutional Neural Networks (CNNs) are considered state-of-the-art in image recognition and offer the ability to provide a prompt and definite diagnosis.  ... 
doi:10.17762/turcomat.v12i3.1581 fatcat:ixaopqqzwrhvvdpsm2zftl77ay

BRepNet: A topological message passing system for solid models [article]

Joseph G. Lambourne, Karl D.D. Willis, Pradeep Kumar Jayaraman, Aditya Sanghi, Peter Meltzer, Hooman Shayani
2021 arXiv   pre-print
BRepNet defines convolutional kernels with respect to oriented coedges in the data structure.  ...  In the neighborhood of each coedge, a small collection of faces, edges and coedges can be identified and patterns in the feature vectors from these entities detected by specific learnable parameters.  ...  As in image convolution, specific entities relative to each coedge map to specific learnable parameters in our convolutional kernels, allowing patterns in the input data to be easily recognized [32, 9  ... 
arXiv:2104.00706v2 fatcat:q2em4c5mafcjnal3d72wqugmru

Automatic Extraction of Green Tide From GF-3 SAR Images Based on Feature Selection and Deep Learning

Haifei Yu, Changying Wang, Jinhua Li, Yi Sui
2021 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
ACKNOWLEDGMENT The GF-3 SAR data were obtained from https://osdds.nsoas.  ...  The authors would like to thank the National Satellite Ocean Application Service (NSOAS) for providing the data free of charge.  ...  CONCLUSION SAR images show high resolution and high timeliness, but in the meantime, they produce massive data and face "data disaster."  ... 
doi:10.1109/jstars.2021.3118374 fatcat:d2w2c5zgjzexbcomi7av74gysq

Off-the-Shelf Deep Features for Saliency Detection

Aymen Azaza, Mehrez Abdellaoui, Ali Douik
2021 SN Computer Science  
Recently, salient object segmentation has introduced the use of object proposals. Object proposal methods provide image segments as proposals which can be used for saliency estimation.  ...  Experimental results shows that we outperform other state-of-the-art methods in PASCAL-S, FT, ECSSD, MSRA-B and ImgSal data sets in terms of F-score, PR curves.  ...  [5] combine low-level bottom-up features with top-down features computed such as face and text. Convolutional Features A typical convolutional network alternates linear and nonlinear layers.  ... 
doi:10.1007/s42979-021-00499-7 fatcat:wgu6awptnfdo5kqrnwh7x5fxny

Robust Underwater Fish Detection Using an Enhanced Convolutional Neural Network

Dipta Gomes, American International University-Bangladesh (AIUB), Dhaka, Bangladesh, A.F.M. Saifuddin Saif
2021 International Journal of Image Graphics and Signal Processing  
The proposed research refines the image enhancement of under-water imagery by proposing an improvement of already existing tools for underwater Object detection.  ...  All this is carried out by overcoming difficulties underwater through a novel technique that can be integrated into an Underwater Autonomous Vehicle and can be classified as robust in nature.  ...  [1] proposed Adaptive Fore-ground Extraction Method using deep Convolutional Neural Network which works very well with dynamic environment and uses a deep Convolution Neural Networks for classifying  ... 
doi:10.5815/ijigsp.2021.03.04 fatcat:sa6bbpulufevfazhl2hqzvwu5q

Video2GIF: Automatic Generation of Animated GIFs from Video [article]

Michael Gygli and Yale Song and Liangliang Cao
2016 arXiv   pre-print
We effectively deal with the noisy web data by proposing a novel adaptive Huber loss in the ranking formulation.  ...  GIFs are short looping video with no sound, and a perfect combination between image and video that really capture our attention.  ...  C3D extends the image-centric network architecture of AlexNet [22] to the video domain by replacing the traditional 2D convolutional layers with a spatio-temporal convolutional layer, and has been shown  ... 
arXiv:1605.04850v1 fatcat:clqzxpc2dzgf5k2gfy7bd6mjry

Operational Neural Networks [article]

Serkan Kiranyaz, Turker Ince, Alexandros Iosifidis, Moncef Gabbouj
2019 arXiv   pre-print
In order to address this drawback and also to accomplish a more generalized model over the convolutional neurons, this study proposes a novel network model, called Operational Neural Networks (ONNs), which  ...  can be heterogeneous and encapsulate neurons with any set of operators to boost diversity and to learn highly complex and multi-modal functions or spaces with minimal network complexity and training data  ...  Moreover, we shall keep the output layer as a convolutional layer whilst optimizing only the two hidden layers by GIS. iii) Scarce Train Data: For the two problems (image denoising and segmentation) with  ... 
arXiv:1902.11106v2 fatcat:frgu77pgfnhrdh656kulzbkgiu

Video2GIF: Automatic Generation of Animated GIFs from Video

Michael Gygli, Yale Song, Liangliang Cao
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
We effectively deal with the noisy web data by proposing a novel adaptive Huber loss in the ranking formulation.  ...  GIFs are short looping video with no sound, and a perfect combination between image and video that really capture our attention.  ...  C3D extends the image-centric network architecture of AlexNet [22] to the video domain by replacing the traditional 2D convolutional layers with a spatio-temporal convolutional layer, and has been shown  ... 
doi:10.1109/cvpr.2016.114 dblp:conf/cvpr/GygliSC16 fatcat:swfrkcb2ujcu5dreot3onopozy

Research on Salient Object Detection using Deep Learning and Segmentation Methods

2019 International journal of recent technology and engineering  
Detecting and segmenting salient objects in natural scenes, often referred to as salient object detection has attracted a lot of interest in computer vision and recently various heuristic computational  ...  It also discusses the open issues in terms of evaluation metrics and dataset bias in model performance and suggests future research directions.  ...  Region-based convolutional networks for accurate object detection and segmentation.  ... 
doi:10.35940/ijrte.b1046.0982s1119 fatcat:6ofq53vb7zhx7boq4ndpraphs4
« Previous Showing results 1 — 15 out of 271 results