Browse
Submit Data
Databases
API
Help

Dataset Information

0 Views

0 Connections

0 Citations

0 Reanalyses

0 Downloads

Omics score: 0

MA-CharNet: Multi-angle fusion character recognition network.

ABSTRACT: Irregular text recognition of natural scene is a challenging task due to large span of character angles and morphological diversity of a word. Recent work first rectifies curved word region, and then employ sequence algorithm to complete the recognition task. However, this strategy largely depends on rectification quality of the text region, and cannot be applied to large difference between tilt angles of character. In this work, a novel anchor-free network structure of rotating character detection is proposed, which includes multiple sub-angle domain branch networks, and the corresponding branch network can be selected adaptively according to character tilt angle. Meanwhile, a curvature Adaptive Text linking method is proposed to connect the discrete strings detected on the two-dimensional plane into words according to people's habits. We achieved state-of-the-art performance on two irregular texts (TotalText, CTW1500), outperforming state-of-the-art by 2.4% and 2.7%, respectively. The experimental results demonstrate the effectiveness of the proposed algorithm.

SUBMITTER: Wang Q

PROVIDER: S-EPMC9423680 | biostudies-literature | 2022

REPOSITORIES: biostudies-literature

ACCESS DATA

Json Xml

Publications

MA-CharNet: Multi-angle fusion character recognition network.

Wang Qingyu Q Liu Jing J Zhu Ziqi Z Deng Chunhua C

PloS one 20220829 8

Irregular text recognition of natural scene is a challenging task due to large span of character angles and morphological diversity of a word. Recent work first rectifies curved word region, and then employ sequence algorithm to complete the recognition task. However, this strategy largely depends on rectification quality of the text region, and cannot be applied to large difference between tilt angles of character. In this work, a novel anchor-free network structure of rotating character detect ...[more]

PMID: 36037174

Similar Datasets

Multi-supervised bidirectional fusion network for road-surface condition recognition.

Project description:Rapid developments in automatic driving technology have given rise to new experiences for passengers. Safety is a main priority in automatic driving. A strong familiarity with road-surface conditions during the day and night is essential to ensuring driving safety. Existing models used for recognizing road-surface conditions lack the required robustness and generalization abilities. Most studies only validated the performance of these models on daylight images. To address this problem, we propose a novel multi-supervised bidirectional fusion network (MBFN) model to detect weather-induced road-surface conditions on the path of automatic vehicles at both daytime and nighttime. We employed ConvNeXt to extract the basic features, which were further processed using a new bidirectional fusion module to create a fused feature. Then, the basic and fused features were concatenated to generate a refined feature with greater discriminative and generalization abilities. Finally, we designed a multi-supervised loss function to train the MBFN model based on the extracted features. Experiments were conducted using two public datasets. The results clearly demonstrated that the MBFN model could classify diverse road-surface conditions, such as dry, wet, and snowy conditions, with a satisfactory accuracy and outperform state-of-the-art baseline models. Notably, the proposed model has multiple variants that could also achieve competitive performances under different road conditions. The code for the MBFN model is shared at https://zenodo.org/badge/latestdoi/607014079.

| S-EPMC10495952 | biostudies-literature

One shot ancient character recognition with siamese similarity network.

Project description:Ancient character recognition is not only important for the study and understanding of ancient history but also has a profound impact on the inheritance and development of national culture. In order to reduce the study of difficult professional knowledge of ancient characters, and meanwhile overcome the lack of data, class imbalance, diversification of glyphs, and open set recognition problems in ancient characters, we propose a Siamese similarity network based on a similarity learning method to directly learn input similarity and then apply the trained model to establish one shot classification task for recognition. Multi-scale fusion backbone structure and embedded structure are proposed in the network to improve the model's ability to extract features. We also propose the soft similarity contrast loss function for the first time, which ensures the optimization of similar images with higher similarity and different classes of images with greater differences while reducing the over-optimization of back-propagation leading to model overfitting. Specially, we propose a cumulative class prototype based on our network to solve the deviation problem of the mean class prototype and obtain a good class representation. Since new ancient characters can still be found in reality, our model has the ability to reject unknown categories while identifying new ones. A large number of experiments show that our proposed method has achieved high-efficiency discriminative performance and obtained the best performance over the methods of traditional deep learning and other classic one-shot learning.

| S-EPMC9436983 | biostudies-literature

Recognition of multi-modal fusion images with irregular interference.

Project description:Recognizing tomatoes fruits based on color images faces two problems: tomato plants have a long fruit bearing period, the colors of fruits on the same plant are different; the growth of tomato plants generally has the problem of occlusion. In this article, we proposed a neural network classification technology to detect maturity (green, orange, red) and occlusion degree for automatic picking function. The depth images (geometric boundary information) information of the fruits were integrated to the original color images (visual boundary information) to facilitate the RGB and depth information fusion into an integrated set of compact features, named RD-SSD, the mAP performance of RD-SSD model in maturity and occlusion degree respectively reached 0.9147.

| S-EPMC9299258 | biostudies-literature

FAN-MCCD: Fast and Accurate Network for Multi-Scale Chinese Character Detection.

Project description:Inaccurate localization due to scale-variation during character detection causes a widespread issue overconfidence in results of the document analysis community, for the most part in historical and handwritten documents. In this work, we explored the performance of a state-of-the-art network with a simple pipeline that fast and accurately predicts handwritten Chinese characters in old documents. In order to adapt to locations of characters with multi-scale more precisely, excluding pre-processing and in-between steps, we utilized a network with multi-scale feature maps. Then, across each feature map, pre-selected boxes of unalike scales and aspect ratios were employed. The last step was to prune the bounding boxes, sending them to non-maximum suppression to yield the final results. Focusing on a well-designed neural network architecture and loss function that presents well-classified examples, we found our experiments on Caoshu, Character, and Src-images datasets demonstrated that detection performance was enhanced for the detection rate (DT), the false positive per character (FPPC), and the F-score in the order of 98.84%, 0.71, and 97.64%, respectively. In comparison with SSD (single-shot detector), the detection performance of a detection rate (DT), the false positive per character (FPPC), and the F-score were 61.12%, 6.12, and 60.33%, respectively.

| S-EPMC8586960 | biostudies-literature

Multi-channel feature fusion attention Dehazing network.

Project description:Haze is a typical weather phenomena that has a significant negative impact on transportation safety, particularly in the port, highways, and airport runway areas. A multi-scale U-shaped dehazing network is proposed in this research, which is based on our multi-channel feature fusion attention structure. With the help of the feature fusion attention techniques, the model can focus on the intriguing locations with higher haze concentration area. In conjunction with UNet, it can achieve multi-scale feature reuse and residual learning, allowing it to fully utilize the feature information of each layer for image restoration. Experimental resulsts show that our technique performs well on a variety of test datasets. On highway data sets, the PSNR / SSIM / L∞ error performance over the novel technique is increased by 0.52% / 0.5% / 30.84%, 4.68% / 0.78% / 26.19% and 13.84% / 9.05% / 55.57% respectively, when compared to DehazeFormer, MIRNetv2, and FSDGN methods. The findings suggest that our proposed method performs better on image dehazing, especially in terms of L∞ error performance.

| S-EPMC10424867 | biostudies-literature

Streamlining data recording through optical character recognition: a prospective multi-center study in intensive care units.

Project description:BackgroundThe manual entry of data into large patient databases requires significant resources and time. It is possible that a system that is enhanced with the technology of optical character recognition (OCR) can facilitate data entry, reduce data entry errors, and decrease the burden on healthcare personnel.MethodsThis was a prospective multi-center observational study across intensive care units (ICU) in 3 countries. Subjects were critically-ill and required invasive mechanical ventilation and extracorporeal life support. Clinical photos from various medical devices were uploaded using an OCR-enhanced case record form. The degree of data completeness, data accuracy, and time saved in entering data were compared with conventional manual data entry.ResultsThe OCR-based system was developed with 868 photos and validated with 469 photos. In independent validation by 8 untrained personnel involving 1018 data points, the overall data completeness was 98.5% (range 98.2-100%), while the overall data accuracy was 96.9% (range 95.3-100%). It significantly reduced data entry time compared to manual entry (mean reduction 43.9% [range 27.0-1.1%]). The average data entry time needed per patient were 3.4 (range 1.2-5.9) minutes with the OCR-based system, compared with 6.0 (range 2.2-8.1) minutes with manual data entry. Users reported high satisfaction with the tool, with an overall recommendation rate of 4.25 ± 1.04 (maximum of 5).ConclusionAn OCR-based data entry system can effectively and efficiently facilitate data entry into clinical databases, making it a promising tool for future clinical data management. Wider uptake of these systems should be encouraged to better understand their strengths and limitations in both clinical and research settings.

| S-EPMC11917072 | biostudies-literature

An Acquisition Method of Agricultural Equipment Roll Angle Based on Multi-Source Information Fusion.

Project description:Accurately obtaining roll angles is one of the key technologies to improve the positioning accuracy and operation quality of agricultural equipment. Given the demand for the acquisition of agricultural equipment roll angles, a roll angle monitoring model based on Kalman filtering and multi-source information fusion was established by using the MTi-300 AHRS inertial sensor (INS) and XW-GI 5630 BeiDou Navigation Satellite System (BDS), which were installed on agricultural equipment. Data of the INS and BDS were fused by MATLAB; then, Kalman filter was used to optimize the data, and the state equation and measurement equation of the integrated system were established. Then, an integrated monitoring terminal man-machine interactive interface was designed on MATLAB GUI, and a roll angle monitoring system based on the INS and BDS was designed and applied into field experiments. The mean absolute error of the integrated monitoring system based on multi-source information fusion during field experiments was 0.72°, which was smaller compared with the mean absolute errors of roll angle monitored by the INS and BDS independently (0.78° and 0.75°, respectively). Thus, the roll angle integrated model improves monitoring precision and underlies future research on navigation and independent operation of agricultural equipment.

| S-EPMC7181222 | biostudies-literature

Data set for Tifinagh handwriting character recognition.

Project description:The Tifinagh alphabet-IRCAM is the official alphabet of the Amazigh language widely used in North Africa [1]. It includes thirty-one basic letter and two letters each composed of a base letter followed by the sign of labialization. Normalized only in 2003 (Unicode) [2], ICRAM-Tifinagh is a young character repertoire. Which needs more work on all levels. In this context we propose a data set for handwritten Tifinagh characters composed of 1376 image; 43 Image For Each character. The dataset can be used to train a Tifinagh character recognition system, or to extract the meaning characteristics of each character.

| S-EPMC4510372 | biostudies-literature

Handwriting Moroccan regions recognition using Tifinagh character.

Project description:The territorial organization of Morocco during administratives division of 2009 is based on 16 regions. In this work we will create a system of recognition of handwritten words (names of regions) using the Amazigh language is an official language by the Moroccan Royal Institute of Amazigh Culture (IRCAM) (2003a) [1] such as this language is slightly treated by researchers in pattern recognition field that is why we decided to study this language (El Kessab et al., 2013 [3]; El Kessab et al., 2014 [4]) that knowing the state make a decision to computerize the various public sectors by this language. In this context we propose a data set for handwritten Tifinagh regions composed of 1600 image (100 Image for each region). The dataset can be used in one hand to test the efficiency of the Tifinagh region recognition system in extraction of characteristics significatives and the correct identification of each region in classification phase in the other hand.

| S-EPMC4783523 | biostudies-literature

A multi-dilated convolution network for speech emotion recognition.

Project description:Speech emotion recognition (SER) is an important application in Affective Computing and Artificial Intelligence. Recently, there has been a significant interest in Deep Neural Networks using speech spectrograms. As the two-dimensional representation of the spectrogram includes more speech characteristics, research interest in convolution neural networks (CNNs) or advanced image recognition models is leveraged to learn deep patterns in a spectrogram to effectively perform SER. Accordingly, in this study, we propose a novel SER model based on the learning of the utterance-level spectrogram. First, we use the Spatial Pyramid Pooling (SPP) strategy to remove the size constraint associated with the CNN-based image recognition task. Then, the SPP layer is deployed to extract both the global-level prominent feature vector and multi-local-level feature vector, followed by an attention model to weigh the feature vectors. Finally, we apply the ArcFace layer, typically used for face recognition, to the SER task, thereby obtaining improved SER performance. Our model achieved an unweighted accuracy of 67.9% on IEMOCAP and 77.6% on EMODB datasets.

| S-EPMC11894186 | biostudies-literature

OmicsDI is part of the ELIXIR infrastructure

OmicsDI is an Elixir interoperability service. Learn more ›

Tweets

OmicsDI Databases

PRIDE
PeptideAtlas
MassIVE
JPOST Repository
Physiome Model Repository

EGA
EVA
ENA
LINCS
PAXDB
Cell Collective

MetaboLights
Metabolomics Workbench
MetabolomeExpress
GNPS
BioModels
FAIRDOMHub

ArrayExpress
dbGaP
ExpressionAtlas
GEO
NODE

Information

Databases
Help
API
Contact us
Code on GitHub
Terms of use
Submit Data