Multilayer Hybrid Deep-Learning Method for Waste Classification and Recycling

Chu, Yi; Huang, Chen; Xie, Xiao-Liang; Tan, Bohai; Kamal, Shyam; Xiong, Xiaogang

doi:10.1155/2018/5060857

Cited by 176 publications

(77 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…One is a single waste image datasets with a total of 2527 images, and another dataset is manually collected by us, with a total of 5000 images. The experiments to compare with counterpart methods are conducted on the same dataset, and the actual performance of the model in different waste categories on these datasets is discussed in detail.…”

Section: Methodsmentioning

confidence: 99%

“…A multilayer hybrid deep‐learning system is proposed to automatically detect recyclable garbage in public areas, which consists of a CNN for extracting image features and a multi‐layer perceptron for consolidating relevant features. This study has achieved 90% accuracy of classification, which significantly outperforms the CNN‐only model.…”

Section: Related Workmentioning

confidence: 99%

“…For example, according to the composition of the waste, there are organic waste, paper, glass, metal, plastic, and so on . According to the disposal method, there are compostable wastes, recyclable wastes, and so on . Also, there are other classification standards, such as construction wastes, e‐wastes, medical wastes, and so on, according to the different industry fields.…”

Section: Waste Image Datasetsmentioning

confidence: 99%

“…In the latest studies and applications, an Support Vector Machine (SVM)‐scale‐invariant feature transform (SIFT) method achieves only 63% classification accuracy for a manually collected single waste image dataset, while the Convolution Neural Network (CNN)+SVM method using ResNet50 as a pretrained model achieves 87% classification accuracy. Garbage sorting devices and mobile applications that have been put into use can achieve about 90% classification accuracy …”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A combination model based on transfer learning for waste classification

Huang

et al. 2020

Concurrency and Computation

View full text Add to dashboard Cite

Summary The increasing amount of solid waste is becoming a significant problem that needs to be addressed urgently. The reliable and accurate classification method is a crucial step in waste disposal because different types of wastes have different disposal ways. The existing waste classification models driven by deep learning are not easy to achieve accurate results and still need to be improved due to the various architecture networks adopted. Their performance on different datasets is varied, and there is also a lack of specific large‐scale datasets for training. We propose a new combination classification model based on three pretrained CNN models (VGG19, DenseNet169, and NASNetLarge) for processing the ImageNet database and achieve high classification accuracy. In our proposed model, the transfer learning model based on each pretrained model is constructed as a candidate classifier, and the optimal output of three candidate classifiers is selected as the final classification result. The experiments based on two waste image datasets demonstrate that the proposed model achieves 96.5% and 94% classification accuracy and outperforms several counterpart methods.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Waste Image Datasetsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A combination model based on transfer learning for waste classification

Huang

et al. 2020

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…In recent years, deep learning has been proven to be the most powerful data representation method (Chao, Zhi, Dong and Liu, 2018;Chu, Huang, Xie, Tan, Kamal and Xiong, 2018;Geng, Zhang, Li, Gu, Liang, Liang, Wang, Wu, Patil and Wang, 2017;Glorot, Bordes and Bengio, 2011;Guo, Liu, Oerlemans, Lao, Wu and Lew, 2016;Hu, Wang, Peng, Qiu, Shi and Liu, 2018;Längkvist, Karlsson and Loutfi, 2014;LeCun, Bengio and Hinton, 2015;Ngiam, Khosla, Kim, Nam, Lee and Ng, 2011;Sadouk, Gadi and Essoufi, 2018;Schmidhuber, 2015;Voulodimos, Doulamis, Bebis and Stathaki, 2018a;Voulodimos, Doulamis, Doulamis and Protopapadakis, 2018b;Wu, Zhai, Li, Cui, Wang and Patil;Zhang, Liang, Li, Fang, Wang, Geng and Wang, 2017;Zhang, Liang, Su, Qu and Wang, 2018a). Deep learning methods learn a neural network of multiple layers to extract the hierarchical patterns from the original data, and provide high-level and abstractive features for the learning problems.…”

Section: Backgroundsmentioning

confidence: 99%

Low-Rank Deep Convolutional Neural Network for Multitask Learning

Shang

Wang³

2019

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

In this paper, we propose a novel multitask learning method based on the deep convolutional network. The proposed deep network has four convolutional layers, three max-pooling layers, and two parallel fully connected layers. To adjust the deep network to multitask learning problem, we propose to learn a low-rank deep network so that the relation among different tasks can be explored. We proposed to minimize the number of independent parameter rows of one fully connected layer to explore the relations among different tasks, which is measured by the nuclear norm of the parameter of one fully connected layer, and seek a low-rank parameter matrix. Meanwhile, we also propose to regularize another fully connected layer by sparsity penalty so that the useful features learned by the lower layers can be selected. The learning problem is solved by an iterative algorithm based on gradient descent and back-propagation algorithms. The proposed algorithm is evaluated over benchmark datasets of multiple face attribute prediction, multitask natural language processing, and joint economics index predictions. The evaluation results show the advantage of the low-rank deep CNN model over multitask problems.

show abstract

A Study on AI‐based Waste Management Strategies for the COVID‐19 Pandemic

et al. 2022

View full text Add to dashboard Cite

COVID-19 has swept across the globe and disrupted all vectors of social life. Every informed measure must be taken to stop its spread, bring down number of new infections and move to normalization of daily life. Contemporary research has not identified waste management as one of the critical transmission vectors for COVID-19 virus. However, most underdeveloped countries are facing problems in waste management processes due to the general inadequacy and inability of waste management. In that context, smart intervention will be needed to contain possibility of the COVID-19 spread due to inadequate waste management. This paper presents a comparative study of the artificial intelligence/ machine learning based techniques, and potential applications in the COVID-19 waste management cycle (WMC). A general integrated solid waste management (ISWM) strategy is mapped for both short-term and long-term goals of COVID-19 WMC, making use of the techniques investigated. By aligning current health/waste-related guidelines from health organizations and governments worldwide and contemporary, relevant research in area, the challenge of COVID-19 waste management and, subsequently, slowing the pandemic down may be assisted.

show abstract

Multilayer Hybrid Deep-Learning Method for Waste Classification and Recycling

Cited by 176 publications

References 13 publications

A combination model based on transfer learning for waste classification

A combination model based on transfer learning for waste classification

Low-Rank Deep Convolutional Neural Network for Multitask Learning

A Study on AI‐based Waste Management Strategies for the COVID‐19 Pandemic

Contact Info

Product

Resources

About