Deep convolutional recurrent neural network with transfer learning for hyperspectral image classification

Abstract

Molecular latent representations, derived from autoencoders (AEs), are widely used for drug or material discovery over past couple of years. In particular, a variety of machine learning methods based on latent representations has shown excellent performance on quantitative structure-activity relationship (QSAR) modelling. However, the sequence feature of them hasn't been considered in most cases. In addition, data scarcity is still the main obstacle for deep learning strategies, especially for biological activity datasets. In this study, we propose the convolutional recurrent neural network and transfer learning (CRNNTL) method inspired by the applications of polyphonic sound detection and electrocardiogram classification. Our strategy takes advantages of both convolutional and recurrent neural networks for feature extraction, as well as the data augmentation method. Herein, CRNNTL is evaluated on 20 benchmark datasets in comparison with baseline methods. In addition, one isomers-based dataset is used to elucidate its ability for both local and global feature extraction. Then, knowledge transfer performance of CRNNTL is tested, especially for small biological activity datasets. Finally, different latent representations from other type of AEs were used for versatility study of our model. The results show the effectiveness of CRNNTL using different latent representation. Moreover, efficient knowledge transfer is achieved to overcome data scarcity considering binding site similarity between different targets.

Deep Relation Network for Hyperspectral Image Few-Shot Classification

“…GRBS, an unsupervised band selection method based on graph representation, can perform better in both accuracy and efficiency. The spatial neighborhood of each pixel is set to 9 × 9 with reference to [25,39,48]. After the above processing, each HSI is transformed into a number of 9 × 9 × 100 data cubes, so as to standardize the data dimensions and optimize the learning process.…”

“…Therefore, we should effectively utilize transferable knowledge in the collected HSI to further classify other new HSI, so as to reduce cost as much as possible. Different HSI contain different types and quantities of ground objects, so it is difficult for the general transfer learning [47,48] to obtain satisfactory classification accuracy with a few labeled sample. According to the idea of meta-learning, the model not only needs to learn transferable knowledge that is conducive to classification but also needs to learn the ability to learn.…”

Fault Diagnosis in Microelectronics Attachment Via Deep Learning Analysis of 3-D Laser Scans

“…Deep Learning (DL) refers to a set of technologies, primarily Artificial Neural Networks (ANN), that has significantly improved state of the art in several tasks such as computer [1] and robotic [2] vision applications, image analysis [3], [4], audio signals [5], [6] and other complex signals [7]. In industrial systems, DL applications focus on inspection and fault diagnosis in industrial shopfloors and manufacturing processes.…”

Transfer Learning for Segmenting Dimensionally Reduced Hyperspectral Images

“…Also, it brings the possibilities of on-board data reduction executed before transferring the acquired data from an imaging satellite. Although there exist works which show the usefulness of transfer learning in HSI segmentation in various fields [14], they are focused on applying this technique to different deep architectures [15]- [19]. To the best of our knowledge, our approach is the first which comprehensively combines effective HSI data reduction and transfer learning.…”