Image robust recognition based on feature-entropy-oriented differential fusion capsule network

In solving the black box attribute problem of neural networks, how to extract feature information in data and generalize inherent features of data are the focus of artificial intelligence research. Aiming at the problem of the weak generalization ability of large image transformation under deep convolutional networks, a new method for image robust recognition based on a feature-entropy-oriented differential fusion capsule network (DFC) is proposed, the core of which is feature entropy approximation. First, convolution feature entropy is introduced as the transformation metric at the feature extraction level, and a convolution difference scale space is constructed using a residual network to approximate the similar entropy. Then, based on this scale feature, convolution feature extraction in a lower scale space is carried out and fused with the last scale feature to form a convolution differential fusion feature. Finally, a capsule network is used to autonomously cluster using dynamic routing to complete the semantic learning of various high-dimensional features, thereby further enhancing the recognition robustness. Experimental results show that feature entropy can effectively evaluate the transformation image recognition effect, and the DFC is effective for robust recognition with large image transformations such as image translation, rotation, and scale transformation.

[1]  Giorgos Tolias,et al.  Fine-Tuning CNN Image Retrieval with No Human Annotation , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Hirokazu Kameoka,et al.  Sequence-to-Sequence Voice Conversion with Similarity Metric Learned Using Generative Adversarial Networks , 2017, INTERSPEECH.

[3]  Dirk Tomandl,et al.  A Modified General Regression Neural Network (MGRNN) with new, efficient training algorithms as a robust 'black box'-tool for data analysis , 2001, Neural Networks.

[4]  Yair Weiss,et al.  Why do deep convolutional networks generalize so poorly to small image transformations? , 2018, J. Mach. Learn. Res..

[5]  Wang Yijie,et al.  Paragraph Vector Representation Based on Word to Vector and CNN Learning , 2018 .

[6]  Xiaohui Zhao,et al.  Points-of-interest recommendation based on convolution matrix factorization , 2017, Applied Intelligence.

[7]  Dorien Herremans,et al.  Singing voice separation using a deep convolutional neural network trained by ideal binary mask and cross entropy , 2018, Neural Computing and Applications.

[8]  Robert P. Anderson,et al.  Maximum entropy modeling of species geographic distributions , 2006 .

[9]  Dacheng Tao,et al.  Transform-Invariant Convolutional Neural Networks for Image Classification and Search , 2016, ACM Multimedia.

[10]  Saeed Shiry Ghidary,et al.  Convolutional Neural Networks for Image Processing: An Application in Robot Vision , 2003, Australian Conference on Artificial Intelligence.

[11]  Yuan Tian,et al.  Application of new advanced CNN structure with adaptive thresholds to color edge detection , 2012 .

[12]  Rangeet Pan,et al.  Static deep neural network analysis for robustness , 2019, ESEC/SIGSOFT FSE.

[13]  Chen Xu,et al.  MS-CapsNet: A Novel Multi-Scale Capsule Network , 2018, IEEE Signal Processing Letters.

[14]  C. Langlotz,et al.  Deep Learning to Classify Radiology Free-Text Reports. , 2017, Radiology.

[15]  Lei Zhang,et al.  Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising , 2016, IEEE Transactions on Image Processing.

[16]  Dean Zhao,et al.  An optimized classification algorithm by BP neural network based on PLS and HCA , 2014, Applied Intelligence.

[17]  Guoping Qiu,et al.  Learning Based Image Transformation Using Convolutional Neural Networks , 2018, IEEE Access.

[18]  Junwei Han,et al.  Learning Rotation-Invariant Convolutional Neural Networks for Object Detection in VHR Optical Remote Sensing Images , 2016, IEEE Transactions on Geoscience and Remote Sensing.