Greetings, fellow Newsletters! May you always be in good health.
This is the IAES Newsletter of the Institute of Advanced Engineering and Science. Today we will share some news about transfer learning. Transfer learning is a machine learning technique that uses knowledge gained from one task for another related task. This reduces the amount of data labeling and resources required and improves model performance and efficiency. Shirahatti et al. (2023) conducted a study aimed at detecting irony and its types in English tweets using a transfer learning approach. The authors proposed a model based on DistilBERT architecture and bidirectional long-short-term memory (Bi-LSTM) network to classify irony in tweets. The proposed system achieved 81% accuracy for non-irony and 66% for irony, 77% recall for non-irony and 72% for irony, and 79% F1 score for non-irony and 69% for irony classes. This research extends the work from binary classification to multiclass classification of irony, providing an overview for future research on different types of irony in tweets.
Fine grained irony classification through transfer learning approach
Abhinandan Shirahatti, Vijay Rajpurohit, Sanjeev Sannakki
Nowadays irony appears to be pervasive in all social media discussion forums and chats, offering further obstacles to sentiment analysis efforts. The aim of the present research work is to detect irony and its types in English tweets We employed a new system for irony detection in English tweets, and we propose a distilled bidirectional encoder representations from transformers (DistilBERT) light transformer model based on the bidirectional encoder representations from transformers (BERT) architecture, this is further strengthened by the use and design of bidirectional long-short term memory (Bi-LSTM) network this configuration minimizes data preprocessing tasks proposed model tests on a SemEval-2018 task 3, 3,834 samples were provided. Experiment results show the proposed system has achieved a precision of 81% for not irony class and 66% for irony class, recall of 77% for not irony and 72% for irony, and F1 score of 79% for not irony and 69% for irony class since researchers have come up with a binary classification model, in this study we have extended our work for multiclass classification of irony. It is significant and will serve as a foundation for future research on different types of irony in tweets.
Further research was conducted by Rattaphun and Songsri-in (2023), who examined the use of transfer learning to classify Thai cultural images. The results show that this is an effective approach, using three pre-trained artificial neural network models.
Thai culture image classification with transfer learning
Munlika Rattaphun, Kritaphat Songsri-in
Classifying images of Thai culture is important for a variety of applications, such as tourism, education, and cultural preservation. However, building a Machine learning model from scratch to classify Thai cultural images can be challenging due to the limited availability of annotated data. In this study, we investigate the use of transfer learning for the task of image classification on a dataset of Thai cultural images. We utilize three popular convolutional neural network models, namely MobileNet, EfficientNet, and residual network (ResNet) as baseline pre-trained models. Their performances were evaluated when they were trained from random initialization, used as a feature extractor, and fully fine-tuned. The results showed that all three models performed better in terms of accuracy and training time when they were used as a feature extractor, with EfficientNet achieving the highest accuracy of 95.87% while maintaining the training time of 24 ms/iteration. To better understand the reasoning behind the predictions made by the models, we deployed the gradient-weighted class activation mapping (Grad-CAM) visualization technique to generate heatmaps that the models attend to when making predictions. Both our quantitative and qualitative experiments demonstrated that transfer learning is an effective approach to image classification on Thai cultural images.
Jain et al. (2023) developed a real-time eyeglass detection framework using facial or eye image features, specifically on nonstandard facial images. They used an artificial neural network based on Inception V3 architecture which gave an accuracy of 99.2% in training and 99.9% in testing.
Real-time eyeglass detection using transfer learning for non-standard facial data
Ritik Jain, Aashi Goyal, Kalaichelvi Venkatesan
The aim of this paper is to build a real-time eyeglass detection framework based on deep features present in facial or ocular images, which serve as a prime factor in forensics analysis, authentication systems and many more. Generally, eyeglass detection methods were executed using cleaned and fine-tuned facial datasets; it resulted in a well-developed model, but the slightest deviation could affect the performance of the model giving poor results on real-time non-standard facial images. Therefore, a robust model is introduced which is trained on custom non-standard facial data. An Inception V3 architecture based pre-trained convolutional neural network (CNN) is used and fine-tuned using model hyper-parameters to achieve a high accuracy and good precision on non-standard facial images in real-time. This resulted in an accuracy score of about 99.2% and 99.9% for training and testing datasets respectively in less amount of time thereby showing the robustness of the model in all conditions.
In another study, Sadanand et al. (2023) tried to develop a real-time mask and social distancing rule violation detection system using a transfer learning approach on MobileNetV2 and YOLOv3 object detection models. This system has high accuracy and can be integrated with IP cameras or surveillance systems.
Social distance and face mask detector system exploiting transfer learning
Vijaya Shetty Sadanand, Keerthi Anand, Pooja Suresh, Punnya Kadyada Arun Kumar, Priyanka Mahabaleshwar
As time advances, the use of deep learning-based object detection algorithms has also evolved leading to developments of new human-computer interactions, facilitating an exploration of various domains. Considering the automated process of detection, systems suitable for detecting violations are developed. One such applications is the social distancing and face mask detectors to control air-borne diseases. The objective of this research is to deploy transfer learning on object detection models for spotting violations in face masks and physical distance rules in real-time. The common drawbacks of existing models are low accuracy and inability to detect in real-time. The MobileNetV2 object detection model and YOLOv3 model with Euclidean distance measure have been used for detection of face mask and physical distancing. A proactive transfer learning approach is used to perform the functionality of face mask classification on the patterns obtained from the social distance detector model. On implementing the application on various surveillance footage, it was observed that the system could classify masked and unmasked faces and if social distancing was maintained or not with accuracies 99% and 94% respectively. The models exhibited high accuracy on testing and the system can be infused with the existing internet protocol (IP) cameras or surveillance systems for real-time surveillance of face masks and physical distancing rules effectively.
Artificial neural network-based plant disease detection modeling with transfer learning has been the focus of research. According to Adebayo et al. (2023), these studies have shown effectiveness and potential to improve model performance and reduce the need for large amounts of training data.
Convolutional neural network-based crop disease detection model using transfer learning approach
Segun Adebayo, Halleluyah Oluwatobi Aworinde, Akinwale O. Akinwunmi, Adebamiji Ayandiji, Awoniran Olalekan Monsir
Crop diseases disrupt the crop’s physiological constitution by affecting the crop’s natural state. The physical recognition of the symptoms of the various diseases has largely been used to diagnose cassava infections. Every disease has a distinct set of symptoms that can be used to identify it. Early detection through physical identification, however, is quite difficult for a vast crop field. The use of electronic tools for illness identification then becomes necessary to promote early disease detection and control. Convolutional neural networks (CNN) were investigated in this study for the electronic identification and categorization of photographs of cassava leaves. For feature extraction and classification, the study used databases of cassava images and a deep convolutional neural network model. The methodology of this study retrained the models’ current weights for visual geometry group (VGG-16), VGG-19, SqueezeNet, and MobileNet. Accuracy, loss, model complexity, and training time were all taken into consideration when evaluating how well the final layer of CNN models performed when trained on the new cassava image datasets.
The above articles are a small part of the research on transfer learning. To get more information, readers can visit the page and read the articles for FREE through the following links: http://iaesprime.com/index.php/csit/, https://ijece.iaescore.com/, dan https://ijeecs.iaescore.com/.
By: I. Busthomi