Neural Architecture Search for Lightweight Neural Network in Food Recognition
Healthy eating is an essential element to prevent obesity that will lead to chronic diseases. Despite numerous efforts to promote the awareness of healthy food consumption, the obesity rate has been increased in the past few years. An automated food recognition system is needed to serve as a fundame...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-05-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/9/11/1245 |
| _version_ | 1797532136931065856 |
|---|---|
| author | Ren Zhang Tan XinYing Chew Khai Wah Khaw |
| author_facet | Ren Zhang Tan XinYing Chew Khai Wah Khaw |
| author_sort | Ren Zhang Tan |
| collection | DOAJ |
| description | Healthy eating is an essential element to prevent obesity that will lead to chronic diseases. Despite numerous efforts to promote the awareness of healthy food consumption, the obesity rate has been increased in the past few years. An automated food recognition system is needed to serve as a fundamental source of information for promoting a balanced diet and assisting users to understand their meal consumption. In this paper, we propose a novel Lightweight Neural Architecture Search (LNAS) model to self-generate a thin Convolutional Neural Network (CNN) that can be executed on mobile devices with limited processing power. LNAS has a sophisticated search space and modern search strategy to design a child model with reinforcement learning. Extensive experiments have been conducted to evaluate the model generated by LNAS, namely LNAS-NET. The experimental result shows that the proposed LNAS-NET outperformed the state-of-the-art lightweight models in terms of training speed and accuracy metric. Those experiments indicate the effectiveness of LNAS without sacrificing the model performance. It provides a good direction to move toward the era of AutoML and mobile-friendly neural model design. |
| first_indexed | 2024-03-10T10:54:50Z |
| format | Article |
| id | doaj.art-a4ead076990b409580fbad4a311daa80 |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-10T10:54:50Z |
| publishDate | 2021-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-a4ead076990b409580fbad4a311daa802023-11-21T21:56:44ZengMDPI AGMathematics2227-73902021-05-01911124510.3390/math9111245Neural Architecture Search for Lightweight Neural Network in Food RecognitionRen Zhang Tan0XinYing Chew1Khai Wah Khaw2School of Computer Sciences, Universiti Sains Malaysia, Gelugor 11800, Pulau Pinang, MalaysiaSchool of Computer Sciences, Universiti Sains Malaysia, Gelugor 11800, Pulau Pinang, MalaysiaSchool of Management, Universiti Sains Malaysia, Gelugor 11800, Pulau Pinang, MalaysiaHealthy eating is an essential element to prevent obesity that will lead to chronic diseases. Despite numerous efforts to promote the awareness of healthy food consumption, the obesity rate has been increased in the past few years. An automated food recognition system is needed to serve as a fundamental source of information for promoting a balanced diet and assisting users to understand their meal consumption. In this paper, we propose a novel Lightweight Neural Architecture Search (LNAS) model to self-generate a thin Convolutional Neural Network (CNN) that can be executed on mobile devices with limited processing power. LNAS has a sophisticated search space and modern search strategy to design a child model with reinforcement learning. Extensive experiments have been conducted to evaluate the model generated by LNAS, namely LNAS-NET. The experimental result shows that the proposed LNAS-NET outperformed the state-of-the-art lightweight models in terms of training speed and accuracy metric. Those experiments indicate the effectiveness of LNAS without sacrificing the model performance. It provides a good direction to move toward the era of AutoML and mobile-friendly neural model design.https://www.mdpi.com/2227-7390/9/11/1245deep learningreinforcement learningConvolutional Neural NetworkNeural Architecture Search |
| spellingShingle | Ren Zhang Tan XinYing Chew Khai Wah Khaw Neural Architecture Search for Lightweight Neural Network in Food Recognition Mathematics deep learning reinforcement learning Convolutional Neural Network Neural Architecture Search |
| title | Neural Architecture Search for Lightweight Neural Network in Food Recognition |
| title_full | Neural Architecture Search for Lightweight Neural Network in Food Recognition |
| title_fullStr | Neural Architecture Search for Lightweight Neural Network in Food Recognition |
| title_full_unstemmed | Neural Architecture Search for Lightweight Neural Network in Food Recognition |
| title_short | Neural Architecture Search for Lightweight Neural Network in Food Recognition |
| title_sort | neural architecture search for lightweight neural network in food recognition |
| topic | deep learning reinforcement learning Convolutional Neural Network Neural Architecture Search |
| url | https://www.mdpi.com/2227-7390/9/11/1245 |
| work_keys_str_mv | AT renzhangtan neuralarchitecturesearchforlightweightneuralnetworkinfoodrecognition AT xinyingchew neuralarchitecturesearchforlightweightneuralnetworkinfoodrecognition AT khaiwahkhaw neuralarchitecturesearchforlightweightneuralnetworkinfoodrecognition |