scholarly journals GI-SleepNet: A Highly Versatile Image-Based Sleep Classification Using a Deep Learning Algorithm

2021 ◽  
Vol 3 (4) ◽  
pp. 581-597
Author(s):  
Tianxiang Gao ◽  
Jiayi Li ◽  
Yuji Watanabe ◽  
Chijung Hung ◽  
Akihiro Yamanaka ◽  
...  
Keyword(s):  
Deep Learning ◽  
Learning Algorithm ◽  
Sleep Stage ◽  
Image Data ◽  
Sleep Research ◽  
Generative Adversarial Network ◽  
Adversarial Network ◽  
Deep Learning Algorithm ◽  
Technical Requirements ◽  
Sleep Stage Classification

Sleep-stage classification is essential for sleep research. Various automatic judgment programs, including deep learning algorithms using artificial intelligence (AI), have been developed, but have limitations with regard to data format compatibility, human interpretability, cost, and technical requirements. We developed a novel program called GI-SleepNet, generative adversarial network (GAN)-assisted image-based sleep staging for mice that is accurate, versatile, compact, and easy to use. In this program, electroencephalogram and electromyography data are first visualized as images, and then classified into three stages (wake, NREM, and REM) by a supervised image learning algorithm. To increase its accuracy, we adopted GAN and artificially generated fake REM sleep data to equalize the number of stages. This resulted in improved accuracy, and as little as one mouse’s data yielded significant accuracy. Due to its image-based nature, the program is easy to apply to data of different formats, different species of animals, and even outside sleep research. Image data can be easily understood; thus, confirmation by experts is easily obtained, even when there are prediction anomalies. As deep learning in image processing is one of the leading fields in AI, numerous algorithms are also available.

scholarly journals GI-SleepNet: A Highly Versatile Image-based Sleep Classification using a Deep Learning Algorithm

2021 ◽  
Author(s):  
Tian Xiang Gao ◽  
Jia Yi Li ◽  
Yuji Watanabe ◽  
Chi Jung Hung ◽  
Akihiro Yamanaka ◽  
...  
Keyword(s):  
Deep Learning ◽  
Learning Algorithm ◽  
Sleep Stage ◽  
Image Data ◽  
Sleep Research ◽  
Generative Adversarial Network ◽  
Adversarial Network ◽  
Deep Learning Algorithm ◽  
Technical Requirements ◽  
Sleep Stage Classification

Abstract Sleep-stage classification is essential for sleep research. Various automatic judgment programs including deep learning algorithms using artificial intelligence (AI) have been developed, but with limitations in data format compatibility, human interpretability, cost, and technical requirements. We developed a novel program called GI-SleepNet, generative adversarial network (GAN)-assisted image-based sleep staging for mice that is accurate, versatile, compact, and easy to use. In this program, electroencephalogram and electromyography data are first visualized as images and then classified into three stages (wake, NREM, and REM) by a supervised image learning algorithm. To increase the accuracy, we adopted GAN and artificially generated fake REM sleep data to equalize the number of stages. This resulted in improved accuracy, and as few as one mouse data yielded significant accuracy. Because of its image-based nature, it is easy to apply to data of different formats, of different species of animals, and even outside of sleep research. Image data can be easily understood by humans, thus especially confirmation by experts is easy when there are some anomalies of prediction. Because deep learning of images is one of the leading fields in AI, numerous algorithms are also available.

scholarly journals Reinforcement based Multi-Model Deep Learning Algorithm for Classification

2020 ◽  
Vol 9 (2) ◽  
pp. 1704-1710
Keyword(s):  
Deep Learning ◽  
Maximum Entropy ◽  
Learning Algorithm ◽  
Dimensional Space ◽  
Image Data ◽  
Space Policy ◽  
Model Classification ◽  
Deep Learning Algorithm ◽  
Robust Model ◽  
Neural Network Algorithm

Nowadays researchers are focused on processing the multi-media data for classifying the queries of end users by using search engines. The hybrid combination of a powerful classifier and deep feature extractor are used to develop a robust model, which is performed in a high dimensional space. In this research, a three different types of algorithms are combined to attain a stochastic belief space policy, where these algorithms include generative adversary modelling, maximum entropy Reinforcement Learning (RL) and belief space planning which leads to develop a multi-model classification algorithm. In the simulation framework, different adversarial behaviours are used to minimize the agent's action predictability, which has resulted the proposed method to attain robustness, while comparing with unmodelled adversarial strategies. The proposed reinforcement based Deep Learning (DL) algorithm can be used as multi-model classification purpose. The single neural network algorithm can perform the classification on text data and image data. The RL learns the appropriate belief space policy from the feature extracted information of the text and image data, the belief space policy is generated based on the maximum entropy computation

scholarly journals An Automated Heart Rate-based Algorithm for Sleep Stage Classification: Validation Using Conventional PSG and Innovative Wearable ECG Device

2021 ◽  
Author(s):  
Nicolò Pini ◽  
Ju Lynn Ong ◽  
Gizem Yilmaz ◽  
Nicholas I. Y. N. Chee ◽  
Zhao Siting ◽  
...  
Keyword(s):  
New York ◽  
Learning Algorithm ◽  
Sleep Stage ◽  
Cost Effective ◽  
Deep Sleep ◽  
Sleep Staging ◽  
Deep Learning Algorithm ◽  
Best Value ◽  
Sleep Stage Classification

Study Objectives: Validate a HR-based deep-learning algorithm for sleep staging named Neurobit-HRV (Neurobit Inc., New York, USA). Methods: The algorithm can perform classification at 2-levels (Wake; Sleep), 3-levels (Wake; NREM; REM) or 4- levels (Wake; Light; Deep; REM) in 30-second epochs. The algorithm was validated using an open-source dataset of PSG recordings (Physionet CinC dataset, n=994 participants) and a proprietary dataset (Z3Pulse, n=52 participants), composed of HR recordings collected with a chest-worn, wireless sensor. A simultaneous PSG was collected using SOMNOtouch. We evaluated the performance of the models in both datasets using Accuracy (A), Cohen's kappa (K), Sensitivity (SE), Specificity (SP). Results: CinC - The highest value of accuracy was achieved by the 2-levels model (0.8797), while the 3-levels model obtained the best value of K (0.6025). The 4-levels model obtained the lowest SE (0.3812) and the highest SP (0.9744) for the classification of Deep sleep segments. AHI and biological sex did not affect sleep scoring, while a significant decrease of performance by age was reported across the models. Z3Pulse - The highest value of accuracy was achieved by the 2-levels model (0.8812), whereas the 3-levels model obtained the best value of K (0.611). For classification of the sleep states, the lowest SE (0.6163) and the highest SP (0.9606) were obtained for the classification of Deep sleep segment. Conclusions: Results demonstrate the feasibility of accurate HR-based sleep staging. The combination of the illustrated sleep staging algorithm with an inexpensive HR device, provides a cost-effective and non-invasive solution easily deployable in the home.

scholarly journals Speech Dereverberation Using Deep Learning Algorithm

2021 ◽  
pp. 127-133
Author(s):  
Dr. S. Saraswathi ◽  
S. Ramya
Keyword(s):  
Learning Algorithm ◽  
Generative Adversarial Network ◽  
Convolutional Networks ◽  
Speech Dereverberation ◽  
Fully Convolutional Networks ◽  
Adversarial Network ◽  
Recent Success ◽  
Deep Learning Algorithm ◽  
Reverb Challenge ◽  
Short Time

This paper focuses on speech derverberation using a single microphone. We investigate the applicability of fully convolutional networks (FCN) to enhance the speech signal represented by short-time Fourier transform (STFT) images in light of their recent success in many image processing applications. We present two variants: a "U-Net," which is an encoder-decoder network with skip connections, and a generative adversarial network (GAN) with the U-Net as the generator, which produces a more intuitive cost function for training. To assess our method, we used data from the REVERB challenge and compared our results to those of other methods tested under the same conditions. In most cases, we discovered that our method outperforms the competing methods.

scholarly journals Wildfire-Detection Method Using DenseNet and CycleGAN Data Augmentation-Based Remote Camera Imagery

2020 ◽  
Vol 12 (22) ◽  
pp. 3715 ◽  
Author(s):  
Minsoo Park ◽  
Dai Quoc Tran ◽  
Daekyo Jung ◽  
Seunghee Park
Keyword(s):  
Deep Learning ◽  
Data Augmentation ◽  
Model Performance ◽  
Image Data ◽  
Detection Accuracy ◽  
Generative Adversarial Network ◽  
Data Imbalance ◽  
Surveillance Camera ◽  
Convolutional Networks ◽  
Adversarial Network

To minimize the damage caused by wildfires, a deep learning-based wildfire-detection technology that extracts features and patterns from surveillance camera images was developed. However, many studies related to wildfire-image classification based on deep learning have highlighted the problem of data imbalance between wildfire-image data and forest-image data. This data imbalance causes model performance degradation. In this study, wildfire images were generated using a cycle-consistent generative adversarial network (CycleGAN) to eliminate data imbalances. In addition, a densely-connected-convolutional-networks-based (DenseNet-based) framework was proposed and its performance was compared with pre-trained models. While training with a train set containing an image generated by a GAN in the proposed DenseNet-based model, the best performance result value was realized among the models with an accuracy of 98.27% and an F1 score of 98.16, obtained using the test dataset. Finally, this trained model was applied to high-quality drone images of wildfires. The experimental results showed that the proposed framework demonstrated high wildfire-detection accuracy.

scholarly journals Tomographic reconstruction with a generative adversarial network

2020 ◽  
Vol 27 (2) ◽  
pp. 486-493 ◽  
Author(s):  
Xiaogang Yang ◽  
Maik Kahnt ◽  
Dennis Brückner ◽  
Andreas Schropp ◽  
Yakub Fam ◽  
...  
Keyword(s):  
Phase Retrieval ◽  
Learning Algorithm ◽  
Tomographic Reconstruction ◽  
Training Data ◽  
Training Procedure ◽  
Contrast Imaging ◽  
Generative Adversarial Network ◽  
Data Set ◽  
Adversarial Network ◽  
Deep Learning Algorithm

This paper presents a deep learning algorithm for tomographic reconstruction (GANrec). The algorithm uses a generative adversarial network (GAN) to solve the inverse of the Radon transform directly. It works for independent sinograms without additional training steps. The GAN has been developed to fit the input sinogram with the model sinogram generated from the predicted reconstruction. Good quality reconstructions can be obtained during the minimization of the fitting errors. The reconstruction is a self-training procedure based on the physics model, instead of on training data. The algorithm showed significant improvements in the reconstruction accuracy, especially for missing-wedge tomography acquired at less than 180° rotational range. It was also validated by reconstructing a missing-wedge X-ray ptychographic tomography (PXCT) data set of a macroporous zeolite particle, for which only 51 projections over 70° could be collected. The GANrec recovered the 3D pore structure with reasonable quality for further analysis. This reconstruction concept can work universally for most of the ill-posed inverse problems if the forward model is well defined, such as phase retrieval of in-line phase-contrast imaging.

scholarly journals Generative adversarial network (GAN) enabled on-chip contact microscopy

2018 ◽  
Author(s):  
Xiongchao Chen ◽  
Hao Zhang ◽  
Tingting Zhu ◽  
Yao Yao ◽  
Di Jin ◽  
...  
Keyword(s):  
Deep Learning ◽  
Spatial Resolution ◽  
Learning Algorithm ◽  
Field Of View ◽  
Sensor Chip ◽  
Large Field ◽  
Generative Adversarial Network ◽  
Resolution Image ◽  
Adversarial Network ◽  
Contact Imaging

We demonstrate a deep learning based contact imaging on a CMOS chip to achieve ∼1 μm spatial resolution over a large field of view of ∼24 mm2. By using regular LED illumination, we acquire the single lower-resolution image of the objects placed approximate to the sensor with unit fringe magnification. For the raw contact-mode lens-free image, the pixel size of the sensor chip limits the spatial resolution. We apply a generative and adversarial network (GAN), a type of deep learning algorithm, to circumvent this limitation and effectively recover much higher resolution image of the objects, permitting sub-micron spatial resolution to be achieved across the entire sensor chip active area, which is also equivalent to the imaging field-of-view (24 mm2) due to unit magnification. This GAN-contact imaging approach eliminates the need of either lens or multi-frame acquisition, being very handy and cost-effective. We demonstrate the success of this approach by imaging the proliferation dynamics of cells directly cultured on the chip.

scholarly journals Generative adversarial networks for modeling reservoirs with permeability anisotropy

2021 ◽  
Vol 1201 (1) ◽  
pp. 012066
Author(s):  
R Guliev
Keyword(s):  
Deep Learning ◽  
Learning Algorithm ◽  
Oil Field ◽  
Main Element ◽  
Generative Adversarial Networks ◽  
Geological Model ◽  
Generative Adversarial Network ◽  
Field Development ◽  
Adversarial Network ◽  
Permeability Data

Abstract The geological model is a main element in describing the characteristics of hydrocarbon reservoirs. These models are usually obtained using geostatistical modeling techniques. Recently, methods based on deep learning algorithms have begun to be applied as a generator of a geologic models. However, there are still problems with how to assimilate dynamic data to the model. The goal of this work was to develop a deep learning algorithm - generative adversarial network (GAN) and demonstrate the process of generating a synthetic geological model: • Without integrating permeability data into the model • With data assimilation of well permeability data into the model The authors also assessed the possibility of creating a pair of generative-adversarial network-ensemble smoother to improve the closed-loop reservoir management of oil field development.

scholarly journals Labelling of data on fundus color pictures used to train a deep learning model enhances its macular pathology recognition capabilities

2021 ◽  
Author(s):  
KhP Takhchidi ◽  
PV Gliznitsa ◽  
SN Svetozarskiy ◽  
AI Bursov ◽  
KA Shusterzon
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Learning Algorithm ◽  
Image Data ◽  
Diagnostic Methods ◽  
Age Related Macular Degeneration ◽  
Screening Programs ◽  
Deep Learning Algorithm ◽  
Age Related ◽  
Fundus Photographs

Retinal diseases remain one of the leading causes of visual impairments in the world. The development of automated diagnostic methods can improve the efficiency and availability of the macular pathology mass screening programs. The objective of this work was to develop and validate deep learning algorithms detecting macular pathology (age-related macular degeneration, AMD) based on the analysis of color fundus photographs with and without data labeling. We used 1200 color fundus photographs from local databases, including 575 retinal images of AMD patients and 625 pictures of the retina of healthy people. The deep learning algorithm was deployed in the Faster RCNN neural network with ResNet50 for convolution. The process employed the transfer learning method. As a result, in the absence of labeling, the accuracy of the model was unsatisfactory (79%) because the neural network selected the areas of attention incorrectly. Data labeling improved the efficacy of the developed method: with the test dataset, the model determined the areas with informative features adequately, and the classification accuracy reached 96.6%. Thus, image data labeling significantly improves the accuracy of retinal color images recognition by a neural network and enables development and training of effective models with limited datasets.

Sign in / Sign up

Export Citation Format

Share Document