Introducing Urdu Digits Dataset with Demonstration of an Efficient and Robust Noisy Decoder-Based Pseudo Example Generator

Abstract

In the present work, we propose a novel method utilizing only a decoder for generation of pseudo-examples, which has shown great success in image classification tasks. The proposed method is particularly constructive when the data are in a limited quantity used for semi-supervised learning (SSL) or few-shot learning (FSL). While most of the previous works have used an autoencoder to improve the classification performance for SSL, using a single autoencoder may generate confusing pseudo-examples that could degrade the classifier’s performance. On the other hand, various models that utilize encoder–decoder architecture for sample generation can significantly increase computational overhead. To address the issues mentioned above, we propose an efficient means of generating pseudo-examples by using only the generator (decoder) network separately for each class that has shown to be effective for both SSL and FSL. In our approach, the decoder is trained for each class sample using random noise, and multiple samples are generated using the trained decoder. Our generator-based approach outperforms previous state-of-the-art SSL and FSL approaches. In addition, we released the Urdu digits dataset consisting of 10,000 images, including 8000 training and 2000 test images collected through three different methods for purposes of diversity. Furthermore, we explored the effectiveness of our proposed method on the Urdu digits dataset by using both SSL and FSL, which demonstrated improvement of 3.04% and 1.50% in terms of average accuracy, respectively, illustrating the superiority of the proposed method compared to the current state-of-the-art models.

MSCNN-LSTM Model for Predicting Return Loss of the UHF Antenna in HF-UHF RFID Tag Antenna

“…The output gate is composed of the activation functions σ and tanh, whose update equation is ( 6), and finally, the update of h t is completed by Eq. (7).…”

“…Artificial neural network (ANN) is a viable solution to the above problems as it provides low computational cost solutions to complicated issues [5][6][7]. In recent years, related scholars have used ANN to construct the mapping relationship between RFID tag antenna dimensions and EM responses to predict the EM performance of the antenna, which saves time in calculating antenna performance and improves antenna design efficiency.…”

Physics-aware deep learning framework for linear elasticity

“…In recent years, driven by the advancement of bigdata-based architectures (Khan et al, 2022a), deep learning (DL) techniques (LeCun et al, 2015) have shown great promises in computer vision (Voulodimos et al, 2018;Roy and Bhaduri, 2021;Roy et al, 2022c;Roy and Bhaduri, 2022;Roy et al, 2022a), object detection (Zhao et al, 2019;Chandio et al, 2022;Roy et al, 2022b;Singh et al, 2023a), image classification (Rawat and Wang, 2017;Irfan et al, 2021;Jamil et al, 2022;Khan et al, 2022b), damage detection (Guo et al, 2022;Glowacz, 2022Glowacz, , 2021 brain-computer interfaces (Roy, 2022b,a,c;Singh et al, 2023b) and across various scientific applications (Butler et al, 2018;Ching et al, 2018;Bose and Roy, 2022).…”

Multi-classification deep neural networks for identification of fish species using camera captured images

“…Because this makes it possible for individual fish to be more easily identified, it is of greater relevance in terms of directing the expansion of the fish farming industry. At this time, the bulk of solutions for fish individual recognition (FIR) [ 27 ] make use of DL models that are built on the framework for fish recognition. The framework for fish recognition consists of three processes: fish object detection, fish feature extraction, and fish feature comparison.…”

“…Even if the SSD method is fast, a large number of the parameters have to be supplied manually, and the process of debugging is challenging. The YOLO series approach is well suited for identifying individual fish because of its rapid speed [ 27 ], high accuracy [ 28 ], simple debugging, and real-time detection capabilities [ 29 ].…”