CNN Architectures Performance Evaluation for Image Classification of Mosquito in Indonesia

Mosquitoes are the vectors of excessively deadly disease spreading around the world. This type of insect spread across the globe with various genus or species on its specific geographic area. In Indonesia, there are 125 species and 18 genera of mosquito. Some of them are carrying the deadly disease that causes heaps of death in Indonesia. Identifying the mosquito species or genus is the crucial component to battle against the spread of infections caused by mosquito. With the fast development of neural network architecture, the mosquito type can be identified without using a microscope but directly from the image captured by a camera or smartphone camera. By using this, the mosquito kind or genus can be recognized swiftly. In this work, the performance evaluation of Convolutional Neural Network (CNN) architectures was performed to analyze the mosquito image classification task by classifying 200 mosquito images data from the Bing Search Engine, which divided into four genera. From the results yielded, the MobileNetV2 architecture performed the best on the tested metrics. The model gave 80% accuracy on the test set.

[1]  Kota Yoshida,et al.  Vision-Based Classification of Mosquito Species: Comparison of Conventional and Deep Learning Methods , 2019, Applied Sciences.

[2]  Quoc V. Le,et al.  EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks , 2019, ICML.

[3]  Vijay Vasudevan,et al.  Learning Transferable Architectures for Scalable Image Recognition , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[4]  Stephen J. Roberts,et al.  Mosquito Detection with Neural Networks: The Buzz of Deep Learning , 2017, ArXiv.

[5]  Hyun Myung,et al.  A Deep Learning-Based Automatic Mosquito Sensing and Control System for Urban Mosquito Habitats , 2019, Sensors.

[6]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[7]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[8]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[9]  Frank Kirchner,et al.  Application of convolutional neural networks for classification of adult mosquitoes in the field , 2019, PloS one.

[10]  Suane Pires P. da Silva,et al.  Lung Nodule Classification via Deep Transfer Learning in CT Lung Images , 2018, 2018 IEEE 31st International Symposium on Computer-Based Medical Systems (CBMS).

[11]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[12]  Mark Sandler,et al.  MobileNetV2: Inverted Residuals and Linear Bottlenecks , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[13]  Giovanni Benelli,et al.  Biological Control of Mosquito Vectors: Past, Present, and Future , 2016, Insects.