Abstract
Handwriting recognition is an important application in pattern recognition. Because handwritten Tai Le has similar characters and the proportion of characters with similar shapes is relatively high, this paper proposes a wavelet deep convolution (WDC) feature combined with an ensemble deep variational sparse Gaussian process (EDVSGP) offline handwritten Tai Le recognition method. A Tai Le recognition sample database is constructed, 2 and 3 levels of wavelet decomposition prior features are extracted, and then, the wavelet a priori feature is transformed into a WDC feature. To avoid the dimensional disaster caused by feature fusion, a dimensionality reduction method combining LDA and PCA is used to reducing the dimensionality of the fused features without reducing the accuracy rate. Moreover, to better recognize handwritten Tai Le characters, 6 deep variationally sparse Gaussian process (DVSGP) models are integrated as an EDVSGP classification model. Using the handwritten Tai Le character database, the proposed method is superior to the existing methods, with an accuracy, recall and F1-score of 94.15%, 94.15%, and 94.15%, respectively. The universality of the method is verified on a dataset of handwritten Chinese characters with similar shapes (HCFC2021.4) and the Devanagari dataset.
Similar content being viewed by others
Data availability
Enquiries about data availability should be directed to the authors.
References
Bhattacharya N, Roy P, Pal U (2018) Sub-stroke-wise relative feature for online indic handwriting recognition. ACM Trans Asian Low-Resour Lang Inf Process 18(2):1–16
Chen Z, Yin F, Zhang XY, Liu CL (2020) MuLTReNets: multilingual text recognition networks for simultaneous script identification and handwriting recognition. Pattern Recognit 108:107555
Damianou AC, Lawrence ND (2013) Deep Gaussian processes. Computer Science, arXiv:1211.0358
Diaz M, Ferrer MA, Quintana JJ (2019) Anthropomorphic features for on-line signatures. IEEE Trans Patt Anal Mach Intell 41(12):2807–2819
Gholami A, Kwon K, Wu B, Tai Z, Yue X, Jin P, Zhao S, & Keutzer K (2018) SqueezeNext: hardware-aware neural network design. IEEE Comput Soc Conf Comput Vis Pattern Recogn, Workshops pp 1719–1728. https://doi.org/10.1109/CVPRW.2018.00215
Han K, Wang Y, Tian Q, Guo J, Xu C, Xu C (2020) GhostNet: more features from cheap operations. Proc IEEE Comput Soc Conf Comput Vision Patt Recognit 9157333:1577–1586
He K, Zhang X, Ren S, Sun J (2015) Deep residual learning for image recognition. Proc IEEE Comput Soc Conf Comput Vision Pattern Recognit. https://doi.org/10.1109/CVPR.2016.90
Hebbal A, Brevault L, Balesdent M et al (2018) Efficient global optimization using deep gaussian processes. IEEE Congr Evol Comput. https://doi.org/10.1109/CEC.2018.8477946
Hensman J, Matthews AGDG, Filippone M et al (2015) MCMC for variationally sparse Gaussian processes. Adv Neural Inf Proces Syst. https://doi.org/10.5555/2969239.2969423
Howard A, Sandler GM, Chen LC (2019) Searching for MobileNetV3. IEEE Int Conf Comput Vision 9008835:1314–1324
Ioffe S, Szegedy C (2015) Batch normalization: accelerating deep network training by reducing internal covariate shift. Int Conf Mach Learn 119782:448–456
Koriyama T, Kobayashi T (2019) Statistical parametric speech synthesis using deep gaussian processes. IEEE ACM Trans Audio Speech Lang Process 27(5):948–959
Krizhevsky A, Sutskever I, Hinton GE (2012) ImageNet classification with deep convolutional neural networks. Adv Neural Inf Proces Syst 2:1097–1105
Liu CL, Yin F, Wang DH, Wang Q F (2011) CASIA online and offline Chinese handwriting databases, Proc Int Conf Doc Anal Recognit, Beijing, China, September, pp 37–41 https://doi.org/10.1109/ICDAR.2011.17
López-Pérez M, García L, Benítez C, Molina R (2020) A contribution to deep learning approaches for automatic classification of volcano-seismic events: deep gaussian processes. IEEE Trans Geosci Remote Sens. https://doi.org/10.1109/TGRS.2020.3022995
Melnyk P, You ZQ, Li KQA (2020) high-performance CNN method for offline handwritten Chinese character recognition and visualization. Soft Comput 24(11):7977–7987
Mesquita DPP, Freitas LA, Gomes JPP, Mattos CLC (2020) LS-SVR as a bayesian RBF network. IEEE Trans Neural Netw Learn Syst 31(10):4389–4393
Narang SR, Jindal MK (2020) On the recognition of Devanagari ancient handwritten characters using SIFT and Gabor features. Soft Comput 24(22):17279–17289
Nguyen TNA, Phung SL, Bouzerdoum A (2020) Hybrid deep learning-gaussian process network for pedestrian lane detection in unstructured scenes. IEEE Trans Neural Netw Learn Syst 31(12):5324–5338
Noury Z, Rezaei M (2020) Deep-CAPTCHA: a deep learning-based CAPTCHA solver for vulnerability assessment https://arxiv.org/abs/2006.08296
Raj M, Abirami S (2020) Structural representation-based off-line Tamil handwritten character recognition. Soft Comput 24(2):1447–1472
Redmon J, Farhadi A (2018) YOLOv3: an Incremental Improvement arXiv:1804.02767
Sandler M, Howard A, Zhu M, Zhmoginov A, Chen L (2018) MobileNetV2: inverted residuals and linear bottlenecks. Proc IEEE Comput Soc Conf Comput Vision Pattern Recognit 8578572:4510–4520
Sdiri B, Kaaniche M, Cheikh FA, Beghdadi A, Elle OJ (2019) Efficient enhancement of stereo endoscopic images based on joint wavelet decomposition and binocular combination. IEEE Trans Med Imaging 38(1):33–45
Shutin D, Buchgraber T, Kulkarni SR, Poor HV (2011) Fast variational sparse bayesian learning with automatic relevance determination for superimposed signals. IEEE Trans Signal Process 59(12):6257–6261
Simonyan K, Zisserman A (2015) Very deep convolutional networks for large-scale image recognition. Int Conf Learn Represent, ICLR - Conf. Track Proc. arXiv:1409.1556
Szegedy C, Vanhoucke V, Ioffe S, Shlens J, Wojna Z (2015) Rethinking the inception architecture for computer vision. IEEE Comput Soc Conf Comput Vision Pattern Recognit. https://doi.org/10.1109/CVPR.2016.308
Tan M, Le QV (2019) MixConv: mixed depthwise convolutional Kernels. Br Mach Vis Conf. https://doi.org/10.1109/CVPR42600.2020.00165
Tan M, Le QV (2019b) EfficientNet: rethinking model scaling for convolutional neural networks. Int Conf Mach Learn 156104:10691–10700
Titsias MK (2009) Variational learning of inducing variables in sparse Gaussian processes. J Mach Learn Res 5:567–574
Wang J, Sun K, Cheng T, Jiang B, Zhao CD, Liu YD, Mu Y, Tan M, Wang X, Liu W, Xiao B (2020b) Deep high-resolution representation learning for visual recognition. IEEE Trans Pattern Anal Mach Intell 43(10):3349–3364
Wang C, Wu Y, Chen P, Hsieh J and Yeh I (2020a) CSPNet: a new backbone that can enhance learning capability of CNN, IEEE Comput. Soc Conf Comput Vis Pattern Recogn, Workshops pp 1571–1580
Xu TB, Yang PP, Zhang XY, Liu CL (2019) LightweightNet: Toward fast and lightweight convolutional neural networks via architecture distillation. Patt Recognit 8:272–284
Yang Q, Yan P, Zhang Y (2020) Low-dose CT image denoising using a generative adversarial network with wasserstein distance and perceptual loss. IEEE T Med Imaging 37(6):1348–1357
Yu D, Zhu Z, Min J (2020) Multi-scale decomposition enhancement algorithm for surface defect images of si3n4 ceramic bearing balls based on stationary wavelet transform. Adv Appl Ceram 120(1):47–57
Zhang XY, Bengiob Y, Liu CL (2017) Online and offline handwritten Chinese character recognition: a comprehensive study and new benchmark. Pattern Recognit 61:348–360
Zhang XY, Yin F, Zhang YM (2018) Drawing and recognizing chinese characters with recurrent neural network. IEEE Trans Pattern Anal Mach Intell 40(4):849–862
Acknowledgements
This research was supported by a project funded by the National Social Science Fund of China (No. 21VJXG043). All support is gratefully acknowledged.
Funding
This work was supported by the National Social Science Fund of China (No. 21VJXG043).
Author information
Authors and Affiliations
Contributions
HG, YL, JZ, and YS made equal contributions regarding the conception of the work, the experimental work, the data analysis, and writing the paper.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflicts of interest.
Human and animals rights
This article does not contain any studies with human participants or animals performed by any of the authors.
Informed consent
All referred studies are highlighted in the literature review.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Guo, H., Liu, Y., Zhao, J. et al. Offline handwritten Tai Le character recognition using wavelet deep convolution features and ensemble deep variationally sparse Gaussian processes. Soft Comput 27, 12439–12455 (2023). https://doi.org/10.1007/s00500-023-07883-w
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00500-023-07883-w