2,489 Hits in 5.7 sec

Scene labeling with LSTM recurrent neural networks

Wonmin Byeon, Thomas M. Breuel, Federico Raue, Marcus Liwicki
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
In our approach, classification, segmentation, and context integration are all carried out by 2D LSTM networks, allowing texture and spatial model parameters to be learned within a single model.  ...  are commonly used for sequence classification.  ...  In our approach, classification, segmentation, and context integration are all carried out by 2D LSTM networks, allowing texture and spatial model parameters to be learned within a single model.  ... 
doi:10.1109/cvpr.2015.7298977 dblp:conf/cvpr/ByeonBRL15 fatcat:uh6kci6s2vfoxgw3svsm2xiszm

Author index

2021 2021 1st International Conference on Electronic and Electrical Engineering and Intelligent System (ICE3IS)  
of Plasmodium Skizon and 143 Gametocytes Malaria Images Using Deep Learning 1 Forecasting Air Quality Using tnasstve-Scote WSN Based on ConvoLutionaL LSTM Network 80 Automatic Feeding of Laying Hens Based  ...  Extraction and MuLticLass SVM ALgorithm Classification of Indonesian Traditional IB.4 18 Snacks Based on Image Using ConvoLutionaL NeuraL Network (CNN) ALgorithm Cervical Precancerous Classification System  ... 
doi:10.1109/ice3is54102.2021.9649666 fatcat:4bdeqcqbr5fr5py3lqhq73xv6q

Deep fusion of gray level co-occurrence matrices for lung nodule classification [article]

Ahmed Saihood, Hossein Karshenas, AhmadReza Naghsh Nilchi
2022 arXiv   pre-print
A new long-short-term-memory (LSTM) based deep fusion structure, is introduced, where, the texture features computed from lung nodules through new volumetric grey-level-co-occurrence-matrices (GLCM) computations  ...  LSTM fusion structure.  ...  Classification Results The three modes of GLCM are of concern, to extract the texture features and provide them to LSTM neural networks for fusion and classification.  ... 
arXiv:2205.05123v1 fatcat:mpgnycr7mnddpju6o7df2eknh4

Optical Gas Sensing with Liquid Crystal Droplets and Convolutional Neural Networks

José Frazão, Susana I. C. J. Palma, Henrique M. A. Costa, Cláudia Alves, Ana C. A. Roque, Margarida Silveira
2021 Sensors  
Here, we investigate the use of deep convolutional neural networks (CNN) as pattern recognition systems to analyse optical textures dynamics in LC droplets exposed to a set of different VOCs.  ...  With our classification models, we show that a single individual droplet can recognise 11 VOCs with small structural and functional differences (F1-score above 93%).  ...  VOC Recognition For classification, 3 dimensional CNNs (CNN3D) and LSTMs with 2 dimensional CNNs (CNN2D+LSTM) working as feature extractors were used and a softmax was used as output layer.  ... 
doi:10.3390/s21082854 pmid:33919620 pmcid:PMC8073403 fatcat:fvu46xfhgnckjbzmsp5gicm63i

Texture, Morphology, and Statistical Analysis to Differentiate Primary Brain Tumors on Two-Dimensional Magnetic Resonance Imaging Scans Using Artificial Intelligence Techniques

Subrata Bhattacharjee, Deekshitha Prakash, Cho-Hee Kim, Hee-Cheol Kim, Heung-Kook Choi
2022 Healthcare Informatics Research  
Therefore, this study was carried out to analyze the two-dimensional (2D) texture, morphology, and statistical features of brain tumors and to perform a classification using artificial intelligence (AI  ...  T1-weighted magnetic resonance imaging (MRI) 2D scans were used for analysis and classification (multiclass and binary).  ...  However, the LSTM model was used in our research to classify handcrafted (texture, morphological, and statistical) features.  ... 
doi:10.4258/hir.2022.28.1.46 pmid:35172090 pmcid:PMC8850171 fatcat:ojhfcomkgbhtrptwfnvbim4rka

Deep Vibro-Tactile Perception for Simultaneous Texture Identification, Slip Detection, and Speed Estimation

Yerkebulan Massalim, Zhanat Kappassov, Huseyin Atakan Varol
2020 Sensors  
The use of vibrotactile sensors can increase the accuracy of texture recognition and preempt the slippage of a grasped object.  ...  A comparative analysis of convolutional neural networks (CNNs), feed-forward neural networks, and long short-term memory networks confirmed that deep CNNs have a higher generalization accuracy.  ...  (b) LSTM network for motion prediction. (c) CNN network for motion prediction. (d) FNN network for texture recognition. (e) LSTM network for motion prediction.  ... 
doi:10.3390/s20154121 pmid:32722353 fatcat:gbe36pfxqvbejnapx6yk6cwucm

Discriminating Native from Non-Native Speech Using Fusion of Visual Cues

Christos Georgakis, Stavros Petridis, Maja Pantic
2014 Proceedings of the ACM International Conference on Multimedia - MM '14  
Long Short-Term Memory Neural Networks (LSTMs) are employed to model accent-related speech dynamics and yield accent-class predictions.  ...  The best feature combinations achieve classification accuracy of 75%, rendering the proposed method a useful accent classification tool in cases of missing or noisy audio stream.  ...  CLASSIFICATION WITH LSTMS Long Short-Term Memory Neural Networks (LSTMs) [3] are in principle a variant of traditional recurrent neural networks.  ... 
doi:10.1145/2647868.2655026 dblp:conf/mm/GeorgakisPP14 fatcat:o5dai7aq4zentayvezi46jzin4

Prostate Cancer Detection using Deep Learning and Traditional Techniques

Saqib Iqbal, Ghazanfar Farooq Siddiqui, Amjad Rehman, Lal Hussain, Tanzila Saba, Usman Tariq, Adeel Ahmed Abbasi
2021 IEEE Access  
The results were compared with hand-crafted features such as texture, morphology, and gray level co-occurrence matrix (GLCM ) using non-deep learning classifiers such as support vector machine (SVM ) Gaussian  ...  The LSTM deep learning method yields performance with sensitivity (98.33%), specificity (100%), PPV (100%), NPV (99.26%), accuracy (99.48%), MCC (0.9879) and AUC (0.9999), where using Deep learning method  ...  Yet 3D images are rich in information than 2D; however, it is time-consuming and slow, where 2D images fast, generate more accurate results [28] . 2D images have excellent sensitivity and short image  ... 
doi:10.1109/access.2021.3057654 fatcat:yaqveuryivfwzgoxxnrgeq3w6u

A Double-Deep Spatio-Angular Learning Framework for Light Field based Face Recognition [article]

Alireza Sepas-Moghaddam, Mohammad A. Haque, Paulo Lobato Correia, Kamal Nasrollahi, Thomas B. Moeslund, Fernando Pereira
2019 arXiv   pre-print
The proposed double-deep learning framework includes a long short-term memory (LSTM) recurrent network whose inputs are VGG-Face descriptions that are computed using a VGG-Very-Deep-16 convolutional neural  ...  The VGG-16 network uses different face viewpoints rendered from a full light field image, which are organised as a pseudo-video sequence.  ...  The combination of VGG and LSTM networks has recently been used to learn spatio-temporal (video) models for different visual classification and description tasks.  ... 
arXiv:1805.10078v3 fatcat:v5z22bsygzeuhm7egxhmntinhi

Landmarks-assisted Collaborative Deep Framework for Automatic 4D Facial Expression Recognition [article]

Muzammil Behzad, Nhat Vo, Xiaobai Li, Guoying Zhao
2020 arXiv   pre-print
During the training stage, the dynamic images are used to train an end-to-end deep network, while the feature vectors of landmark images are used train a long short-term memory (LSTM) network.  ...  As well, the given 3D landmarks are projected on a 2D plane as binary images and convolutional layers are used to extract sequences of feature vectors for every landmark video.  ...  Finally, we create a long short-term memory (LSTM) network with a sequence input layer, Bi-LSTM layer with 2000 hidden units, 50% dropout layer followed by FC, Softmax and classification layer.  ... 
arXiv:1910.05445v2 fatcat:z7uk5raxdreo5ftw3l3cr7rsge

An Efficient Three-Dimensional Convolutional Neural Network for Inferring Physical Interaction Force from Video

Dongyi Kim, Hyeon Cho, Hochul Shin, Soo-Chul Lim, Wonjun Hwang
2019 Sensors  
In detail, we could predict the interaction force by observing the texture changes of the target object by an external force.  ...  For this purpose, our hypothesis is that a three-dimensional (3D) convolutional neural network (CNN) can be made to predict the physical interaction forces from video images.  ...  A similar approach [29] was applied to 3D classification using 3 × 3 × 3 filters.  ... 
doi:10.3390/s19163579 fatcat:zedf6bjhbjh7lf5zfvplsbombe

Recurrence Plot-Based Approach for Cardiac Arrhythmia Classification Using Inception-ResNet-v2

Hua Zhang, Chengyu Liu, Zhimin Zhang, Yujie Xing, Xinwen Liu, Ruiqing Dong, Yu He, Ling Xia, Feng Liu
2021 Frontiers in Physiology  
Therefore, the promising results indicate that the 2D RP-based method has a high clinical potential for CA classification using fewer lead ECG signals.  ...  During implementation, the optimal leads (lead II and lead aVR) were selected, and then 1D ECG segments were transformed into 2D texture images by the RP approach.  ...  METHODOLOGY In this work, the classification of the CA problem is modeled as a 2D image classification task using RP-based texture images and the Inception-ResNet-v2 architecture.  ... 
doi:10.3389/fphys.2021.648950 pmid:34079470 pmcid:PMC8165394 fatcat:qpsineswfzcc7dbqbxbwsphvhi

Human Action Recognition from Various Data Modalities: A Review [article]

Zehua Sun, Qiuhong Ke, Hossein Rahmani, Mohammed Bennamoun, Gang Wang, Jun Liu
2021 arXiv   pre-print
of useful yet distinct information and have various advantages depending on the application scenarios.  ...  Consequently, lots of existing works have attempted to investigate different types of approaches for HAR using various modalities.  ...  LSTM networks have also been used for HAR using CSI signal [293] , [294] , [295] . Huang et al.  ... 
arXiv:2012.11866v4 fatcat:twjnaur2jzahznci6clkadylay

Classification and phenological staging of crops from in situ image sequences bydeep learning

2022 Turkish Journal of Electrical Engineering and Computer Sciences  
The learning performances improve with the size of the temporal window and the fine-tuning of the deep convolutional neural network used for feature extraction.  ...  The performances achieved with the proposed system are superior to those obtained by applying classical machine learning methods to handcrafted texture and color features.  ...  The LSTM unit The LSTM [11] network unit is a popular variant of the recurrent neural networks that uses state information.  ... 
doi:10.55730/1300-0632.3850 fatcat:633v2cza7jaj7kh7xx665owtae

Exploiting Spatial Structure for Localizing Manipulated Image Regions

Jawadul H. Bappy, Amit K. Roy-Chowdhury, Jason Bunk, Lakshmanan Nataraj, B.S. Manjunath
2017 2017 IEEE International Conference on Computer Vision (ICCV)  
We perform end-to-end training of the network to learn the parameters through back-propagation given groundtruth mask information.  ...  The recent success of the deep learning approaches in different recognition tasks inspires us to develop a high confidence detection framework which can localize manipulated regions in an image.  ...  Finally, we obtain the 2D map with confidence score of each pixel using three consecutive convolutional layers on top of the LSTM network.  ... 
doi:10.1109/iccv.2017.532 dblp:conf/iccv/BappyRBNM17 fatcat:baslmteninhdxe7k7rheyxohpa
« Previous Showing results 1 — 15 out of 2,489 results