Filters








2,746 Hits in 3.3 sec

2D Attentional Irregular Scene Text Recognizer [article]

Pengyuan Lyu, Zhicheng Yang, Xinhang Leng, Xiaojun Wu, Ruiyu Li, Xiaoyong Shen
2019 arXiv   pre-print
Irregular scene text, which has complex layout in 2D space, is challenging to most previous scene text recognizers.  ...  Recently, some irregular scene text recognizers either rectify the irregular text to regular text image with approximate 1D layout or transform the 2D image feature map to 1D feature sequence.  ...  In this paper, we propose a 2D attentional irregular scene text recognizer which transforms the irregular text with 2D layout to character sequence directly.  ... 
arXiv:1906.05708v1 fatcat:6u6o7upenjfj5fklmdae3u75yy

A Holistic Representation Guided Attention Network for Scene Text Recognition [article]

Lu Yang, Fan Dang, Peng Wang, Hui Li, Zhen Li, Yanning Zhang
2021 arXiv   pre-print
Reading irregular scene text of arbitrary shape in natural images is still a challenging problem, despite the progress made recently.  ...  With this simple design, our method achieves state-of-the-art or competitive recognition performance on the evaluated regular and irregular scene text benchmark datasets.  ...  Figure 5 : 5 The comparison of our proposed 2D attention based and the rectification based (MORAN [37] ) irregular text recognizers.  ... 
arXiv:1904.01375v5 fatcat:v6klgi54z5edrjsh5t5o5wyduq

2D Positional Embedding-based Transformer for Scene Text Recognition

Zobeir Raisi, Mohamed A. Naiel, Paul Fieguth, Steven Wardell, John Zelek
2021 Journal of Computational Vision and Imaging Systems  
In this paper, we leverage a Transformer-based architecture for recognizing both regular and irregular text-in-the-wild images.  ...  irregular-text instances due to the loss of spatial information present in the original two-dimensional (2D) images.  ...  As shown in Figure 2 (a), the proposed method recognized correctly all these images that mostly contain irregular text.  ... 
doi:10.15353/jcvis.v6i1.3533 fatcat:q5hid4xetrf7fnmwzeocjapmq4

Show, Attend and Read: A Simple and Strong Baseline for Irregular Text Recognition [article]

Hui Li, Peng Wang, Chunhua Shen, Guyu Zhang
2019 arXiv   pre-print
Recognizing irregular text in natural scene images is challenging due to the large variance in text appearance, such as curvature, orientation and distortion.  ...  Despite its simplicity, the proposed method is robust and achieves state-of-the-art performance on both regular and irregular scene text recognition benchmarks.  ...  (Shi et al. 2018) irregular text recognizers.  ... 
arXiv:1811.00751v2 fatcat:3gazh47ozfdx5g26bnv4vtnguu

Decoupled Attention Network for Text Recognition [article]

Tianwei Wang, Yuanzhi Zhu, Lianwen Jin, Canjie Luo, Xiaoxue Chen, Yaqiang Wu, Qianying Wang, Mingxiang Cai
2019 arXiv   pre-print
Experimental results show that DAN achieves state-of-the-art performance on multiple text recognition tasks, including offline handwritten text recognition and regular/irregular scene text recognition.  ...  Text recognition has attracted considerable research interests because of its various applications. The cutting-edge text recognition methods are based on attention mechanisms.  ...  4), DAN becomes a 2D recognizer and is suitable for irregular text recognition.  ... 
arXiv:1912.10205v1 fatcat:w46ifcgp7bdabpn7zm65u3yvwa

Decoupled Attention Network for Text Recognition

Tianwei Wang, Yuanzhi Zhu, Lianwen Jin, Canjie Luo, Xiaoxue Chen, Yaqiang Wu, Qianying Wang, Mingxiang Cai
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Experimental results show that DAN achieves state-of-the-art performance on multiple text recognition tasks, including offline handwritten text recognition and regular/irregular scene text recognition.  ...  Text recognition has attracted considerable research interests because of its various applications. The cutting-edge text recognition methods are based on attention mechanisms.  ...  4), DAN becomes a 2D recognizer and is suitable for irregular text recognition.  ... 
doi:10.1609/aaai.v34i07.6903 fatcat:xfkh52nvzfb6rkx3mpowwew2kq

Scene Text Recognition via Transformer [article]

Xinjie Feng, Hongxun Yao, Yuankai Qi, Jun Zhang, Shengping Zhang
2020 arXiv   pre-print
What all we need is the spatial attention. We therefore propose a simple but extremely effective scene text recognition method based on transformer [50].  ...  Scene text recognition with arbitrary shape is very challenging due to large variations in text shapes, fonts, colors, backgrounds, etc.  ...  The proposed 2D attentional scheme transforms the irregular text with the 2D layout to character sequence directly.  ... 
arXiv:2003.08077v4 fatcat:wb2svxx66rgqjnhfqclbpms73e

On Recognizing Texts of Arbitrary Shapes with 2D Self-Attention [article]

Junyeop Lee, Sungrae Park, Jeonghun Baek, Seong Joon Oh, Seonghyeon Kim, Hwalsuk Lee
2019 arXiv   pre-print
SATRN utilizes the self-attention mechanism to describe two-dimensional (2D) spatial dependencies of characters in a scene text image.  ...  Scene text recognition (STR) is the task of recognizing character sequences in natural scenes.  ...  SATRN thus models long-range dependencies spanning 2D space, a feature necessary for recognizing texts of irregular geometry. interpreting texts with arbitrary shapes, which are important challenges in  ... 
arXiv:1910.04396v1 fatcat:3uzeunopavhjbh6p2y7xhi5tgq

Towards End-to-End Text Spotting in Natural Scenes [article]

Peng Wang, Hui Li, Chunhua Shen
2021 arXiv   pre-print
By employing the 2D attention model in word recognition, the irregularity of text can be robustly addressed.  ...  Text spotting in natural scene images is of great importance for many image understanding tasks. It includes two sub-tasks: text detection and recognition.  ...  The use of 2D attention mechanism enables our model detect and recognize curved text with a single forward pass in cluttered natural scene images. 3 )Fig. 9 - 39 Fig. 9 -Visualization of 2D attention  ... 
arXiv:1906.06013v6 fatcat:6yijskgdkjd2tposdbgw7xtfvq

Learning to Read Irregular Text with Attention Mechanisms

Xiao Yang, Dafang He, Zihan Zhou, Daniel Kifer, C. Lee Giles
2017 Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence  
We present a robust end-to-end neural-based model to attentively recognize text in natural images.  ...  loss that provides guidance to the training of an attention model.  ...  Recognizing irregular text in natural scene is addressed by first learning text-specific visual representations, then decoding the learned representations into a character sequence via an attention-based  ... 
doi:10.24963/ijcai.2017/458 dblp:conf/ijcai/YangHZKG17 fatcat:xrbs4lz2yvf4jhqdednr4a52oa

GTC: Guided Training of CTC towards Efficient and Accurate Scene Text Recognition

Wenyang Hu, Xiaocong Cai, Jun Hou, Shuai Yi, Zhiping Lin
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Connectionist Temporal Classification (CTC) and attention mechanism are two main approaches used in recent scene text recognition works.  ...  With the benefit of guided training, CTC model achieves robust and accurate prediction for both regular and irregular scene text while maintaining a fast inference speed.  ...  Irregular Scene Text Recognition Recognizing irregular scene text has attracted increasing attention in recent years, as it is a more challenging problem.  ... 
doi:10.1609/aaai.v34i07.6735 fatcat:5liq5l3p2ffkrpujk4n6dzc3fe

GTC: Guided Training of CTC Towards Efficient and Accurate Scene Text Recognition [article]

Wenyang Hu, Xiaocong Cai, Jun Hou, Shuai Yi, Zhiping Lin
2020 arXiv   pre-print
Connectionist Temporal Classification (CTC) and attention mechanism are two main approaches used in recent scene text recognition works.  ...  With the benefit of guided training, CTC model achieves robust and accurate prediction for both regular and irregular scene text while maintaining a fast inference speed.  ...  Irregular Scene Text Recognition Recognizing irregular scene text has attracted increasing attention in recent years, as it is a more challenging problem.  ... 
arXiv:2002.01276v1 fatcat:qonprtztnfd3pnryf43pbepyhe

Text Detection and Recognition in the Wild: A Review [article]

Zobeir Raisi, Mohamed A. Naiel, Paul Fieguth, Steven Wardell, John Zelek
2020 arXiv   pre-print
Second, identifying several existing challenges for detecting or recognizing text in the wild images, namely, in-plane-rotation, multi-oriented and multi-resolution text, perspective distortion, illumination  ...  The current state-of-the-art scene text detection and/or recognition methods have exploited the witnessed advancement in deep learning architectures and reported a superior accuracy on benchmark datasets  ...  [151] proposed a framework called Character Attention FCN (CA-FCN) , which models the irregular scene text recognition problem in a 2D space instead of the 1D space as well.  ... 
arXiv:2006.04305v2 fatcat:paccfprli5arbj4ggfx5z3hrve

NRTR: A No-Recurrence Sequence-to-Sequence Model For Scene Text Recognition [article]

Fenfen Sheng, Zhineng Chen, Bo Xu
2019 arXiv   pre-print
NRTR follows the encoder-decoder paradigm, where the encoder uses stacked self-attention to extract image features, and the decoder applies stacked self-attention to recognize texts based on encoder output  ...  Considering scene image has large variation in text and background, we further design a modality-transform block to effectively transform 2D input images to 1D sequences, combined with the encoder to extract  ...  Besides, unlike [9] that uses 1D sentence as input, scene text recognizer receives 2D images with large variation in scales/aspect ratios and backgrounds.  ... 
arXiv:1806.00926v2 fatcat:z26tjlmyuzgcdmp5lsmrf5zz6i

MASTER: Multi-Aspect Non-local Network for Scene Text Recognition [article]

Ning Lu, Wenwen Yu, Xianbiao Qi, Yihao Chen, Ping Gong, Rong Xiao
2020 arXiv   pre-print
Attention based scene text recognizers have gained huge success, which leverage a more compact intermediate representations to learn 1d- or 2d- attention by a RNN-based encoder-decoder architecture.  ...  Extensive experiments on various benchmarks demonstrate the superior performance of our MASTER on both regular and irregular scene text.  ...  Existing irregular scene text recognizers can be divided into three categories: rectification based, multi-direction encoding based, and attention-based approaches.  ... 
arXiv:1910.02562v2 fatcat:oqucwqm7zndifozplgexbko6x4
« Previous Showing results 1 — 15 out of 2,746 results