Semi-supervised learning with convolutional neural networks for UAV images automatic recognition

Abstract The annotation of large datasets is an issue whose challenge increases as the number of labeled samples available to train the classifier reduces in comparison to the amount of unlabeled data. In this context, semi-supervised learning methods aim at discovering and propagating labels to unsupervised samples, such that their correct labeling can improve the classification performance. Our proposal makes use of semi-supervised methodologies to classify an unlabeled training set that is used to train a Convolution Neural Network using different training strategies. The proposed approach is experimentally validated for soybean leaf and herbivorous pest identification using images captured by Unmanned Aerial Vehicles and can support specialists and farmers in the pest control management in soybean fields, especially when they have a limited amount of labeled samples.

[1]  Erfu Yang,et al.  A Novel Active Semisupervised Convolutional Neural Network Algorithm for SAR Image Recognition , 2017, Comput. Intell. Neurosci..

[2]  João Paulo Papa,et al.  Petroleum well drilling monitoring through cutting image analysis and artificial intelligence techniques , 2011, Eng. Appl. Artif. Intell..

[3]  João Paulo Papa,et al.  Automatic classification of plant electrophysiological responses to environmental stimuli using machine learning and interval arithmetic , 2018, Comput. Electron. Agric..

[4]  Yoshua Bengio,et al.  Greedy Layer-Wise Training of Deep Networks , 2006, NIPS.

[5]  Yann LeCun,et al.  Convolutional networks and applications in vision , 2010, Proceedings of 2010 IEEE International Symposium on Circuits and Systems.

[6]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[7]  Yoshua Bengio,et al.  Gated Feedback Recurrent Neural Networks , 2015, ICML.

[8]  Hao Wu,et al.  Semi-Supervised Deep Learning Using Pseudo Labels for Hyperspectral Image Classification , 2018, IEEE Transactions on Image Processing.

[9]  Pascal Vincent,et al.  Representation Learning: A Review and New Perspectives , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[10]  Dit-Yan Yeung,et al.  Semi-Supervised Discriminant Analysis using robust path-based similarity , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[11]  Jiebo Luo,et al.  Weakly Semi-Supervised Deep Learning for Multi-Label Image Annotation , 2015, IEEE Transactions on Big Data.

[12]  Dit-Yan Yeung,et al.  Robust path-based spectral clustering , 2008, Pattern Recognit..

[13]  Jason Weston,et al.  Deep learning via semi-supervised embedding , 2008, ICML '08.

[14]  João Paulo Papa,et al.  Improving semi-supervised learning through optimum connectivity , 2016, Pattern Recognit..

[15]  Mikhail Belkin,et al.  Manifold Regularization: A Geometric Framework for Learning from Labeled and Unlabeled Examples , 2006, J. Mach. Learn. Res..

[16]  Pascal Fua,et al.  SLIC Superpixels Compared to State-of-the-Art Superpixel Methods , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[17]  J. P. Papa,et al.  Fast and Robust Object Tracking Using Image Foresting Transform , 2009, 2009 16th International Conference on Systems, Signals and Image Processing.

[18]  João Paulo Papa,et al.  Efficient supervised optimum-path forest classification for large datasets , 2012, Pattern Recognit..

[19]  Shih-Wei Lin,et al.  Particle swarm optimization for parameter determination and feature selection of support vector machines , 2008, Expert Syst. Appl..

[20]  Qiang Yang,et al.  A Survey on Transfer Learning , 2010, IEEE Transactions on Knowledge and Data Engineering.

[21]  João Paulo Papa,et al.  Aquatic weed automatic classification using machine learning techniques , 2012 .

[22]  Ronald M. Summers,et al.  Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning , 2016, IEEE Transactions on Medical Imaging.

[23]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[24]  Nelson D. A. Mascarenhas,et al.  Fast and accurate holistic face recognition using Optimum-Path Forest , 2009, 2009 16th International Conference on Digital Signal Processing.

[25]  João Paulo Papa,et al.  Spoken emotion recognition through optimum-path forest classification using glottal features , 2010, Comput. Speech Lang..

[26]  Thorsten Joachims,et al.  Transductive Inference for Text Classification using Support Vector Machines , 1999, ICML.

[27]  Cheng Wu,et al.  Semi-Supervised and Unsupervised Extreme Learning Machines , 2014, IEEE Transactions on Cybernetics.

[28]  João Paulo Papa,et al.  Robust Pruning of Training Patterns for Optimum-Path Forest Classification Applied to Satellite-Based Rainfall Occurrence Estimation , 2010, IEEE Geoscience and Remote Sensing Letters.

[29]  João Paulo Papa,et al.  Robust and fast Vowel Recognition Using Optimum-Path Forest , 2010, 2010 IEEE International Conference on Acoustics, Speech and Signal Processing.

[30]  Kurt Driessens,et al.  Using Weighted Nearest Neighbor to Benefit from Unlabeled Data , 2006, PAKDD.

[31]  M. Friedman.  A Comparison of Alternative Tests of Significance for the Problem of $m$ Rankings , 1940 .

[32]  Alexandros Iosifidis,et al.  Regularized extreme learning machine for multi-view semi-supervised action recognition , 2014, Neurocomputing.

[33]  Jurandy Almeida,et al.  A Multiple Labeling-Based Optimum-Path Forest for Video Content Classification , 2013, 2013 XXVI Conference on Graphics, Patterns and Images.