596 Hits in 3.6 sec

Deep Convolutional Ranking for Multilabel Image Annotation [article]

Yunchao Gong, Yangqing Jia, Thomas Leung, Alexander Toshev, Sergey Ioffe
2014 arXiv   pre-print
While existing work usually use conventional visual features for multilabel annotation, features based on Deep Neural Networks have shown potential to significantly boost performance.  ...  Multilabel image annotation is one of the most important challenges in computer vision with many real-world applications.  ...  Discussion and Future Work In this work, we proposed to use ranking to train deep convolutional neural networks for multilabel image annotation problems.  ... 
arXiv:1312.4894v2 fatcat:x2hzjzoidbhubovmpce4pos4g4

Automatic image annotation method based on a convolutional neural network with threshold optimization

Jianfang Cao, Aidi Zhao, Zibang Zhang
2020 PLoS ONE  
During the annotation process, the multilabel annotation for the image to be annotated is realized by loading the optimal model and the optimal threshold.  ...  In this study, a convolutional neural network with threshold optimization (CNN-THOP) is proposed to solve the issue of overlabeling or downlabeling arising during the multilabel image annotation process  ...  [19] developed an end-to-end automatic image annotation model based on a deep convolutional neural network (E2E-DCNN) and multilabel data augmentation.  ... 
doi:10.1371/journal.pone.0238956 pmid:32966319 pmcid:PMC7511011 fatcat:tl47iyp7hbhudnb4x5t7m56yni

Multilabel convolution neural network for facial expression recognition and ordinal intensity estimation

Olufisayo Ekundayo, Serestina Viriri
2021 PeerJ Computer Science  
This work presents a Multilabel Convolution Neural Network (ML-CNN)-based model, which could simultaneously recognise emotion and provide ordinal metrics as the intensity estimation of the emotion.  ...  Multilabel ARAM (MLARAM).  ...  ML-CNN model combines multilabel problem transformation techniques with CNN algorithm as a deep learning technique for the multilabel classification task.  ... 
doi:10.7717/peerj-cs.736 pmid:34909462 pmcid:PMC8641570 fatcat:jsv3tmfejfadxdzc7umux4qal4

Global Context-Based Multilevel Feature Fusion Networks for Multilabel Remote Sensing Image Scene Classification

Xin Wang, Lin Duan, Chen Ning
2021 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
Different from the traditional remote sensing (RS) scene classification which uses a single scene label to holistically annotate an image, multilabel RS image classification uses a series of object labels  ...  Experimental results demonstrate that the proposed method is superior to some popular networks for multilabel RS image scene classification.  ...  For example, Tan et al. [4] proposed a low rank representation based algorithm for RS image MLC. Zhou et al. [5] introduced a multiinstance multilabel learning framework for scene understanding.  ... 
doi:10.1109/jstars.2021.3122464 fatcat:iumvszpu25dszdgp5fkpxqqudq

CM-supplement network model for reducing the memory consumption during multilabel image annotation

Jianfang Cao, Lichao Chen, Chenyan Wu, Zibang Zhang, Yan Chai Hum
2020 PLoS ONE  
In the field of image retrieval, image auto-annotation remains a basic and challenging task.  ...  Targeting the drawbacks of the low accuracy rate and high memory resource consumption of current multilabel annotation methods, this study proposed a CM-supplement network model.  ...  Furthermore, deep-level CNN-based discriminative models have also made certain achievements in multilabel image auto-annotation.  ... 
doi:10.1371/journal.pone.0234014 pmid:32479515 fatcat:cg3dowr3sfcnlhqiylgwm4sxx4

Gated recurrent units and temporal convolutional network for multilabel classification [article]

Loris Nanni, Alessandra Lumini, Alessandro Manfe, Riccardo Rampon, Sheryl Brahnam, Giorgio Venturin
2021 arXiv   pre-print
This work proposes a new ensemble method for managing multilabel classification: the core of the proposed approach combines a set of gated recurrent units and temporal convolutional neural networks trained  ...  Multilabel learning tackles the problem of associating a sample with multiple class labels.  ...  Zisserman, "Very deep convolutional networks for large-scale image recognition," Cornell University, arXiv:1409.1556v6 2014. [35] C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, and Z.  ... 
arXiv:2110.04414v2 fatcat:gyqka2igufhvpau7arqciqtbju

CNN-RNN: A Unified Framework for Multi-label Image Classification

Jiang Wang, Yi Yang, Junhua Mao, Zhiheng Huang, Chang Huang, Wei Xu
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
While deep convolutional neural networks (CNNs) have shown a great success in single-label image classification, it is important to note that real world images generally contain multiple labels, which  ...  Traditional approaches to multi-label image classification learn independent classifiers for each category and employ ranking or thresholding on the classification results.  ...  Deep convolutional ranking [9] optimizes a top-k ranking objective, which assigns smaller weights to the loss if the positive label.  ... 
doi:10.1109/cvpr.2016.251 dblp:conf/cvpr/WangYMHHX16 fatcat:s2tgck7esbbl5nmfxluycs6cea

End-to-end Binary Representation Learning via Direct Binary Embedding [article]

Liu Liu, Alireza Rahimpour, Ali Taalimi, Hairong Qi
2017 arXiv   pre-print
Extensive experiments demonstrate the significant superiority of DBE over state-of-the-art methods on tasks of natural object recognition, image retrieval and image annotation.  ...  By employing the deep residual network (ResNet) as DCNN component, DBE captures rich semantics from images.  ...  multilabel image annotation.  ... 
arXiv:1703.04960v2 fatcat:476wazltkfdjdnw2whqo2c3ybe

Discriminative Cross-View Binary Representation Learning [article]

Liu Liu, Hairong Qi
2018 arXiv   pre-print
First, it uses convolutional neural network (CNN) based nonlinear hashing functions and multilabel classification for both images and texts simultaneously.  ...  In addition, DCVH can provide competitive performance for image annotation/tagging.  ...  Explicitly, a deep convolutional neural network [9] projects images into lowerdimensional latent feature space; for texts, DCVH uses pretrained GloVe [24] vectors to represent words in the texts.  ... 
arXiv:1804.01233v1 fatcat:5vvdhxgshzbsxpr57mbl4la6i4

Learning Image Conditioned Label Space for Multilabel Classification [article]

Yi-Nan Li, Mei-Chen Yeh
2018 arXiv   pre-print
Inspired by the great success from deep convolutional neural networks (CNNs) for single-label visual-semantic embedding, we exploit extending these models for multilabel images.  ...  Specifically, we propose an image-dependent ranking model, which returns a ranked list of labels according to its relevance to the input image.  ...  simple and compact, yet it is more powerful than many existing deep learningbased models for multilabel classification.  ... 
arXiv:1802.07460v1 fatcat:jg7bo4jumzfldmitkhcozgivou

MultiScene: A Large-Scale Dataset and Benchmark for Multiscene Recognition in Single Aerial Images

Yuansheng Hua, Lichao Mou, Pu Jin, Xiao Xiang Zhu
2021 IEEE Transactions on Geoscience and Remote Sensing  
With it, we can develop and evaluate deep networks for multiscene recognition using clean data.  ...  Index Terms-Convolutional neural network (CNN), crowdsourced annotations, large-scale aerial image dataset, learning from noisy labels, multiscene recognition in single images.  ...  For the latter, we leverage the same test set but train deep neural networks on the other 93 000 images with only crowdsourced annotations. 2) Evaluation: For a comprehensive evaluation, we measure the  ... 
doi:10.1109/tgrs.2021.3110314 fatcat:fbhxrmgv6fhmlht5wct76ayhxy

Fusing Multilabel Deep Networks for Facial Action Unit Detection

Mina Bishay, Ioannis Patras
2017 2017 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017)  
At the core of our approach is a novel dynamic adaptation of the Deep Network cost function so as to deal with the data imbalances that are inherent in multilabel classification problems -this allows crossdatabase  ...  We show the benefits of the proposed training approach and how different architectures are more suitable for particular AUs.  ...  At the first stage, two Convolutional Neural Networks (CNNs) are used for learning deep appearance features based on cropped face images, and two Multi-Layer Perceptrons (MLPs) for learning distinctive  ... 
doi:10.1109/fg.2017.86 dblp:conf/fgr/BishayP17 fatcat:hdy5obzl2balflk72qo2rubw6u

Deep semantic ranking based hashing for multi-label image retrieval

Fang Zhao, Yongzhen Huang, Liang Wang, Tieniu Tan
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
Here we propose a deep semantic ranking based method for learning hash functions that preserve multilevel semantic similarity between multilabel images.  ...  In our approach, deep convolutional neural network is incorporated into hash functions to jointly learn feature representations and mappings from them to hash codes, which avoids the limitation of semantic  ...  between multilabel images.  ... 
doi:10.1109/cvpr.2015.7298763 dblp:conf/cvpr/ZhaoHWT15 fatcat:t2u7qqcqgzg7fpd5jg46tikh5i

Automatic image annotation based on deep learning models: a systematic review and future challenges

Myasar Mundher Adnan, Mohd Shafry Mohd Rahim, Amjad Rehman, Zahid Mehmood, Tanzila Saba, Rizwan Ali Naqvi
2021 IEEE Access  
To address a fixed number of labels appearing during the multilabel image annotation process and label annotation according to the ranking function.  ...  [42] This article proposed an end-to-end deep learning framework for multilabel annotation of RS images that exploits dual-level semantic concepts.  ... 
doi:10.1109/access.2021.3068897 fatcat:yqhu53zrlbe7jpntvd4sal6yqu

A Retrospective Comparison of Deep Learning to Manual Annotations for Optic Disc and Optic Cup Segmentation in Fundus Photographs

Huazhu Fu, Fei Li, Yanwu Xu, Jingan Liao, Jian Xiong, Jianbing Shen, Jiang Liu, Xiulan Zhang, for iChallenge-GON study group
2020 Translational Vision Science & Technology  
Optic disc (OD) and optic cup (OC) segmentation are fundamental for fundus image analysis.  ...  A deep learning system for OD and OC segmentation was developed.  ...  deep network in the cropped OD image.  ... 
doi:10.1167/tvst.9.2.33 pmid:32832206 pmcid:PMC7414704 fatcat:ppv6pa47nnfspbqhaxpbfetzxe
« Previous Showing results 1 — 15 out of 596 results