Low-res MobileNet: An efficient lightweight network for low-resolution image classification in resource-constrained scenarios

Yuan, Haiying; Cheng, Junpeng; Wu, Yanrui; Zeng, Zhiyong

doi:10.1007/s11042-022-13157-8

Cited by 10 publications

(4 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Reduce computing complexity and memory requirements: In resource-constrained scenarios, lightweight model structures can be developed to reduce computational complexity and memory requirements [ 60 ]. In addition, model compression and distillation techniques can be used to maintain performance while reducing resource requirements.…”

Section: Discuss and Future Workmentioning

confidence: 99%

ThyroidNet: A Deep Learning Network for Localization and Classification of Thyroid Nodules

Chen,

Pan

et al. 2024

CMES

View full text Add to dashboard Cite

Aim This study aims to establish an artificial intelligence model, ThyroidNet, to diagnose thyroid nodules using deep learning techniques accurately. Methods A novel method, ThyroidNet, is introduced and evaluated based on deep learning for the localization and classification of thyroid nodules. First, we propose the multitask TransUnet, which combines the TransUnet encoder and decoder with multitask learning. Second, we propose the DualLoss function, tailored to the thyroid nodule localization and classification tasks. It balances the learning of the localization and classification tasks to help improve the model’s generalization ability. Third, we introduce strategies for augmenting the data. Finally, we submit a novel deep learning model, ThyroidNet, to accurately detect thyroid nodules. Results ThyroidNet was evaluated on private datasets and was comparable to other existing methods, including U-Net and TransUnet. Experimental results show that ThyroidNet outperformed these methods in localizing and classifying thyroid nodules. It achieved improved accuracy of 3.9% and 1.5%, respectively. Conclusion ThyroidNet significantly improves the clinical diagnosis of thyroid nodules and supports medical image analysis tasks. Future research directions include optimization of the model structure, expansion of the dataset size, reduction of computational complexity and memory requirements, and exploration of additional applications of ThyroidNet in medical image analysis.

show abstract

Section: Discuss and Future Workmentioning

confidence: 99%

ThyroidNet: A Deep Learning Network for Localization and Classification of Thyroid Nodules

Chen,

Pan

et al. 2024

CMES

View full text Add to dashboard Cite

show abstract

“…The effectiveness of lightweight neural networks in homogeneous low-resolution face recognition has been studied, 10 and it was shown that lightweight face models can achieve similar or even better performance than complex models on Low-resolution face recognition work. 39 Recently, there are more architectures reported in low-resolution face recognition areas, such as the Tailored MobileNet V2 network architecture for low-resolution features extraction, 40 LRFRHop 7 for resource-constrained environments, and multiscale parallel deep CNN (mpdCNN) 41 for face recognition in low-resolution images. It is worth noting that the LRFRHop model was designed based on the Successive Subspace Learning principle.…”

Section: Related Workmentioning

confidence: 99%

A homogeneous low-resolution face recognition method using correlation features at the edge

Zhao,

Nagothu,

Chen

2024

Sensors and Systems for Space Applications XVII

View full text Add to dashboard Cite

Face recognition technology has been well investigated in past decades and widely deployed in many realworld applications. However, low-resolution face recognition is still a challenging task in resource-constrained edge computing environment like the Internet of Video Things (IoVT) applications. For instance, low-resolution images are common in surveillance video streams, in which the rare information, variable angles, and light conditions create difficulties for recognition tasks. To address these problems, we optimized the correlation feature face recognition (CoFFaR) method and conducted experimental studies in two data preparation modes, symmetric and exhaustive arranging. The experimental results show that the CoFFaR method achieved an accuracy rate of over 82.56%, and the two-dimensional (2D) feature points after dimension reduction are uniformly distributed in a diagonal pattern. The analysis leads to the conclusion that the data augmentation advantage brought by the method of exhaustive arranging data preparation can effectively improve the performance, and the constraints by making the feature vector closer to its clustering center have no apparent improvement in the accuracy of the model identification.

show abstract

“…Since all the mudra images were resized to low dimensions for the recognition, the MobileNet which was built by training relatively low resolution images targeting the mobile camera vision applications, showed relatively higher performance for mudra classification. The issues of low resolution-based image classification by conventional CNNs have been discussed in the recent article by Yuan et al 62 MobileNet has been reported to be effective for the object and image classification tasks used in vision systems in vehicles wherein the low resolution images are used for recognition with reduced latency. 63 For the experiments carried out in the paper, all originally acquired images were resized to 64 × 64 and were further down scaled to 32 × 32 to feed as input to the CNNs.…”

Section: Combination Of Eigenmudra and Raw Image-based Cnn Classifica...mentioning

confidence: 99%

Automatic one-hand gesture (mudra) identification in bharatanatyam using eigenmudra projections and convolutional neural networks

2023

View full text Add to dashboard Cite

Mudras in traditional Indian dance forms convey meaningful information when performed by an artist. The subtle changes between the different mudras in a dance form render automatic identification challenging as compared to conventional hand gesture identification, where the gestures are uniquely distinct from each other. Therefore, the objective of this study is to build a classifier model for the identification of the asamyukta mudra of bharatanatyam, one of the most popular classical dance forms in India. The first part of the paper provides a comprehensive review of the issues present in bharatanatyam mudra identification and the various studies conducted on the automatic classification of mudras. Based on this review, we observe that the unavailability of a large mudra corpus is a major challenge in mudra identification. Therefore, the second part of the paper focuses on the development of a relatively large database of mudra images consisting of 29 asamyukta mudras prevalent in bharatanatyam, which is obtained by incorporating different variabilities, such as subject, artist type (amateur or professional), and orientation. The mudra image database so developed is made available for academic research purposes. The final part of this paper describes the development of a convolutional neural network (CNN)-based automatic mudra identification system. Multistyle training of mudra classes on a conventional CNN showed a 92% correct identification rate. Based on the "eigenface" projection used in face recognition, "eigenmudras" projections of mudra images are proposed for improving the CNN-based mudra identification. Although the CNNs trained on the eigenmudra-projected images provide nearly equal identification rates as that obtained using the CNNs trained on raw mudra grayscale images, both models provide complementary mudra class information. The presence of complementary class information is confirmed by the improvement in the mudra identification performance when the CNN models trained from the raw mudra and eigenmudra-projected images are combined by computing the average of the scores obtained in the final softmax layers of both models. The same trend of improved mudra identification is observed upon combination of the average score level of VGG19 CNN models of the raw mudra images and corresponding eigenmudra-projected images.

show abstract

Low-res MobileNet: An efficient lightweight network for low-resolution image classification in resource-constrained scenarios

Cited by 10 publications

References 26 publications

ThyroidNet: A Deep Learning Network for Localization and Classification of Thyroid Nodules

ThyroidNet: A Deep Learning Network for Localization and Classification of Thyroid Nodules

A homogeneous low-resolution face recognition method using correlation features at the edge

Automatic one-hand gesture (mudra) identification in bharatanatyam using eigenmudra projections and convolutional neural networks

Contact Info

Product

Resources

About