Combining Convolutional Neural Network With Recursive Neural Network for Blood Cell Image Classification

Liang, Gaobo; Hong, Huichao; Xie, Weifang; Zheng, Lixin

doi:10.1109/access.2018.2846685

Cited by 227 publications

(121 citation statements)

References 31 publications

Supporting

Mentioning

102

Contrasting

Unclassified

Order By: Relevance

“…Their experiment used 3,000 epochs to fine-tune all layers, trained on 11,200 samples, and evaluated 1,244 WBCs. Liang et al [20] propose a framework that combines the CNN with the recurrent neural network (RNN) to deeply understand the image content and learn the structured features of images. The best performance of this framework reaches 90.97% when combined with Xception and long short-term memory (LSTM).…”

Section: Related Workmentioning

confidence: 99%

Improved White Blood Cells Classification based on Pre-trained Deep Learning Models

Mohamed

El-Behaidy

Khoriba

et al. 2020

J. commun. softw. syst. (Online)

View full text Add to dashboard Cite

Leukocytes, or white blood cells (WBCs), are microscopic organisms that fight against infectious disease, bacteria, viruses and others. The manual method to classify and count WBCs is tedious, time-consuming and may have inaccurate results, whereas the automated methods are costly. This research aims to automatically identify and classify WBCs in a microscopic image into four types with higher accuracy. BCCD is the used dataset in this study, which is a scaled-down blood cell detection dataset. BCCD is firstly preprocessed by passing through various processes such as segmentation and augmentation; then, it is passed to the proposed model. Our model combines the advantage of deep models in automatically extracting features with the higher classification accuracy of traditional machine learning classifiers. The proposed model consists of two main stages: a shallow tuning pre-trained model and a traditional machine learning classifier on top of it. In this study, ten different pretrained models with six types of machine learning are used. Moreover, the fully connected network (FCN) of pre-trained models is used as a baseline classifier for comparison. The evaluation process shows that the hybrid of MobileNet-224 as a feature extractor and logistic regression as classifier has a higher rank-1 accuracy of 97.03%. Besides, the proposed hybrid model outperformed the baseline FCN by 25.78% on average.

show abstract

Section: Related Workmentioning

confidence: 99%

Improved White Blood Cells Classification based on Pre-trained Deep Learning Models

Mohamed

El-Behaidy

Khoriba

et al. 2020

J. commun. softw. syst. (Online)

View full text Add to dashboard Cite

show abstract

“…Alex Net was being used before in the previous models. Deep learning performs well in image categorization tasks [2][3][4][5] which can be noted from its recent improvements. The proposed method uses the RCNN-RNN algorithm and tensor learning for categorizing real-world object images.…”

Section: Literature Surveymentioning

confidence: 99%

“…The pre-trained CNN [9] model on the Image Net dataset and maintain its weight parameters are used. [2] The RCNN is frozen at the beginning and the RNN layer is loaded with the pre-trained data set. [2] The attributes from the RNN are extracted and by merging the RCNN and RNN we get the final image classification.…”

Section: Literature Surveymentioning

confidence: 99%

“…[2] The RCNN is frozen at the beginning and the RNN layer is loaded with the pre-trained data set. [2] The attributes from the RNN are extracted and by merging the RCNN and RNN we get the final image classification. [2] The parameters of the training data are constantly adjusted in order to improve the precision of RNN.…”

Section: Literature Surveymentioning

confidence: 99%

“…[2] The attributes from the RNN are extracted and by merging the RCNN and RNN we get the final image classification. [2] The parameters of the training data are constantly adjusted in order to improve the precision of RNN. In the end, all the layers are removed and the data is used as feedback for the CNN and RNN model.…”

Section: Literature Surveymentioning

confidence: 99%

See 2 more Smart Citations

Real-Time Object Recognition using Region based Convolution Neural Network and Recursive Neural Network

Priyatharshini*,

Ram. A.S,

Sundar

et al. 2019

IJRTE

View full text Add to dashboard Cite

The recognition of real-world objects demands the recognition and characterization of digital image samples. Automated methods for the detection and recognition of entity types have many significant commercial and industrial applications. While deep convolution neural networks (CNN) and machine learning (ML) concepts have contributed to the classification of globe items, they cannot fully scale the reliance of powerful GPUs to classify the key attributes of images. By using a Recurrent Neural Network (RNN) we tend to resolve the issue arisen in the previous systems. In particular, a hybrid approach using R-CNN and RNN has been proposed that improve the accuracy of object recognition and learn structured image attributes and begin image analysis. Specifically, we applied the transfer learning approach to pass the load parameters which were pre-trained on the Image web dataset to the RNN portion and follow a custom loss feature for the model to train and test more rapidly with precise weight parameters. Experimental results show that in comparison to CNN models like Resent, origin V3, etc., our proposed model achieved improved accuracy in categorizing universe pictures.

show abstract

A hybrid approach for efficient multi‐classification of white blood cells based on transfer learning techniques and traditional machine learning methods

Cengil

Çınar

Yıldırım

2021

Concurrency and Computation

View full text Add to dashboard Cite

Diagnosed blood-related diseases include the identification of blood samples taken from the patient. Therefore, the classification of white blood cells, also known as leukocytes, is substantial for differentiating leukemia and blood diseases. In this article, we aimed to classify four different types of white blood cells. Two different transfer learning methods are applied to achieve this goal. In the first method, AlexNet, ResNet18, and GoogleNet architectures are retrained with the fine-tuning method using the dataset we have provided from the Kaggle and then given to classify on the softmax and SVM methods. With this method, the hybrid architecture where the ResNet18 is used with SVM achieves 99.83%.In the second method, feature transfer, the same architectures are implemented for feature extractors. First, extracted features of architectures are given to a variety of classifiers. Second, the features are taken from the architectures, and concatenated as pairs and triples, used to obtain 4 different feature sets consisting of 2000 and 3000 features. After that, these features are also subjected to the same classifiers. Finally, results are revealed by using the same classifiers after the PCA algorithm. The results illustrate that the proposed methods significantly contribute to white blood cell multi-classification.

show abstract

Combining Convolutional Neural Network With Recursive Neural Network for Blood Cell Image Classification

Cited by 227 publications

References 31 publications

Improved White Blood Cells Classification based on Pre-trained Deep Learning Models

Improved White Blood Cells Classification based on Pre-trained Deep Learning Models

Real-Time Object Recognition using Region based Convolution Neural Network and Recursive Neural Network

A hybrid approach for efficient multi‐classification of white blood cells based on transfer learning techniques and traditional machine learning methods

Contact Info

Product

Resources

About