scholarly journals Improvement of Heterogeneous Transfer Learning Efficiency by Using Hebbian Learning Principle

2020 ◽  
Vol 10 (16) ◽  
pp. 5631 ◽  
Author(s):  
Arjun Magotra ◽  
Juntae Kim
Keyword(s):  
Image Classification ◽  
Transfer Learning ◽  
Hebbian Learning ◽  
Learning Rule ◽  
Transfer Of Knowledge ◽  
Learning Efficiency ◽  
Simultaneous Activation ◽  
Classification Of Images ◽  
The Individual ◽  
Heterogeneous Source

Transfer learning algorithms have been widely studied for machine learning in recent times. In particular, in image recognition and classification tasks, transfer learning has shown significant benefits, and is getting plenty of attention in the research community. While performing a transfer of knowledge among source and target tasks, homogeneous dataset is not always available, and heterogeneous dataset can be chosen in certain circumstances. In this article, we propose a way of improving transfer learning efficiency, in case of a heterogeneous source and target, by using the Hebbian learning principle, called Hebbian transfer learning (HTL). In computer vision, biologically motivated approaches such as Hebbian learning represent associative learning, where simultaneous activation of brain cells positively affect the increase in synaptic connection strength between the individual cells. The discriminative nature of learning for the search of features in the task of image classification fits well to the techniques, such as the Hebbian learning rule—neurons that fire together wire together. The deep learning models, such as convolutional neural networks (CNN), are widely used for image classification. In transfer learning, for such models, the connection weights of the learned model should adapt to new target dataset with minimum effort. The discriminative learning rule, such as Hebbian learning, can improve performance of learning by quickly adapting to discriminate between different classes defined by target task. We apply the Hebbian principle as synaptic plasticity in transfer learning for classification of images using a heterogeneous source-target dataset, and compare results with the standard transfer learning case. Experimental results using CIFAR-10 (Canadian Institute for Advanced Research) and CIFAR-100 datasets with various combinations show that the proposed HTL algorithm can improve the performance of transfer learning, especially in the case of a heterogeneous source and target dataset.

scholarly journals Image Classification Using Transfer Learning and Deep Learning

2021 ◽  
Vol 10 (9) ◽  
pp. 25394-25398
Author(s):  
Chitra Desai
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Image Classification ◽  
Transfer Learning ◽  
Network Model ◽  
Neural Network Model ◽  
Large Scale ◽  
Deep Neural Network ◽  
Visual Recognition ◽  
Classification Of Images

Deep learning models have demonstrated improved efficacy in image classification since the ImageNet Large Scale Visual Recognition Challenge started since 2010. Classification of images has further augmented in the field of computer vision with the dawn of transfer learning. To train a model on huge dataset demands huge computational resources and add a lot of cost to learning. Transfer learning allows to reduce on cost of learning and also help avoid reinventing the wheel. There are several pretrained models like VGG16, VGG19, ResNet50, Inceptionv3, EfficientNet etc which are widely used.   This paper demonstrates image classification using pretrained deep neural network model VGG16 which is trained on images from ImageNet dataset. After obtaining the convolutional base model, a new deep neural network model is built on top of it for image classification based on fully connected network. This classifier will use features extracted from the convolutional base model.

scholarly journals Pretraining Convolutional Neural Networks for Mudstone Petrographic Thin-Section Image Classification

Geosciences ◽  
2021 ◽  
Vol 11 (8) ◽  
pp. 336
Author(s):  
Rafael Pires de Lima ◽  
David Duarte
Keyword(s):  
Neural Networks ◽  
Image Classification ◽  
Transfer Learning ◽  
Thin Section ◽  
Convolutional Neural Networks ◽  
Primary Task ◽  
Secondary Tasks ◽  
The Difference ◽  
Classification Of Images ◽  
Petrographic Thin Section

Convolutional neural networks (CNN) are currently the most widely used tool for the classification of images, especially if such images have large within- and small between- group variance. Thus, one of the main factors driving the development of CNN models is the creation of large, labelled computer vision datasets, some containing millions of images. Thanks to transfer learning, a technique that modifies a model trained on a primary task to execute a secondary task, the adaptation of CNN models trained on such large datasets has rapidly gained popularity in many fields of science, geosciences included. However, the trade-off between two main components of the transfer learning methodology for geoscience images is still unclear: the difference between the datasets used in the primary and secondary tasks; and the amount of available data for the primary task itself. We evaluate the performance of CNN models pretrained with different types of image datasets—specifically, dermatology, histology, and raw food—that are fine-tuned to the task of petrographic thin-section image classification. Results show that CNN models pretrained on ImageNet achieve higher accuracy due to the larger number of samples, as well as a larger variability in the samples in ImageNet compared to the other datasets evaluated.

scholarly journals Neuromodulated Dopamine Plastic Networks for Heterogeneous Transfer Learning with Hebbian Principle

Symmetry ◽  
2021 ◽  
Vol 13 (8) ◽  
pp. 1344
Author(s):  
Arjun Magotra ◽  
Juntae Kim
Keyword(s):  
Neural Networks ◽  
Transfer Learning ◽  
Hebbian Learning ◽  
Synaptic Connection ◽  
Novel Approach ◽  
Learning And Adaptation ◽  
Algorithmic Technique ◽  
Learning Principles ◽  
The Brain ◽  
Heterogeneous Source

The plastic modifications in synaptic connectivity is primarily from changes triggered by neuromodulated dopamine signals. These activities are controlled by neuromodulation, which is itself under the control of the brain. The subjective brain’s self-modifying abilities play an essential role in learning and adaptation. The artificial neural networks with neuromodulated plasticity are used to implement transfer learning in the image classification domain. In particular, this has application in image detection, image segmentation, and transfer of learning parameters with significant results. This paper proposes a novel approach to enhance transfer learning accuracy in a heterogeneous source and target, using the neuromodulation of the Hebbian learning principle, called NDHTL (Neuromodulated Dopamine Hebbian Transfer Learning). Neuromodulation of plasticity offers a powerful new technique with applications in training neural networks implementing asymmetric backpropagation using Hebbian principles in transfer learning motivated CNNs (Convolutional neural networks). Biologically motivated concomitant learning, where connected brain cells activate positively, enhances the synaptic connection strength between the network neurons. Using the NDHTL algorithm, the percentage of change of the plasticity between the neurons of the CNN layer is directly managed by the dopamine signal’s value. The discriminative nature of transfer learning fits well with the technique. The learned model’s connection weights must adapt to unseen target datasets with the least cost and effort in transfer learning. Using distinctive learning principles such as dopamine Hebbian learning in transfer learning for asymmetric gradient weights update is a novel approach. The paper emphasizes the NDHTL algorithmic technique as synaptic plasticity controlled by dopamine signals in transfer learning to classify images using source-target datasets. The standard transfer learning using gradient backpropagation is a symmetric framework. Experimental results using CIFAR-10 and CIFAR-100 datasets show that the proposed NDHTL algorithm can enhance transfer learning efficiency compared to existing methods.

scholarly journals Landslide Mapping Using Multiscale LiDAR Digital Elevation Models

2020 ◽  
Vol 26 (4) ◽  
pp. 405-425
Author(s):  
Javed Miandad ◽  
Margaret M. Darrow ◽  
Michael D. Hendricks ◽  
Ronald P. Daanen
Keyword(s):  
Image Classification ◽  
Vegetation Index ◽  
Confusion Matrix ◽  
Object Oriented ◽  
Support Vector ◽  
Classification Approach ◽  
Resolution Image ◽  
Kappa Value ◽  
The Individual ◽  
Segmentation Image

ABSTRACT This study presents a new methodology to identify landslide and landslide-susceptible locations in Interior Alaska using only geomorphic properties from light detection and ranging (LiDAR) derivatives (i.e., slope, profile curvature, and roughness) and the normalized difference vegetation index (NDVI), focusing on the effect of different resolutions of LiDAR images. We developed a semi-automated object-oriented image classification approach in ArcGIS 10.5 and prepared a landslide inventory from visual observation of hillshade images. The multistage work flow included combining derivatives from 1-, 2.5-, and 5-m-resolution LiDAR, image segmentation, image classification using a support vector machine classifier, and image generalization to clean false positives. We assessed classification accuracy by generating confusion matrix tables. Analysis of the results indicated that LiDAR image scale played an important role in the classification, and the use of NDVI generated better results. Overall, the LiDAR 5-m-resolution image with NDVI generated the best results with a kappa value of 0.55 and an overall accuracy of 83 percent. The LiDAR 1-m-resolution image with NDVI generated the highest producer accuracy of 73 percent in identifying landslide locations. We produced a combined overlay map by summing the individual classified maps that was able to delineate landslide objects better than the individual maps. The combined classified map from 1-, 2.5-, and 5-m-resolution LiDAR with NDVI generated producer accuracies of 60, 80, and 86 percent and user accuracies of 39, 51, and 98 percent for landslide, landslide-susceptible, and stable locations, respectively, with an overall accuracy of 84 percent and a kappa value of 0.58. This semi-automated object-oriented image classification approach demonstrated potential as a viable tool with further refinement and/or in combination with additional data sources.

scholarly journals Image Classification and Prediction using Transfer Learning in Colab Notebook

2021 ◽  
Author(s):  
J Praveen Gujjar ◽  
R Prasanna Kumar H ◽  
Niranjan N. Chiplunkar
Keyword(s):  
Image Classification ◽  
Transfer Learning
Sign in / Sign up

Export Citation Format

Share Document