OBC306: A Large-Scale Oracle Bone Character Recognition Dataset

Huang, Shuangping; Wang, Haobin; Liu, Yongge; Xiaosong, Shi; Jin, Lianwen

doi:10.1109/icdar.2019.00114

Cited by 35 publications

(27 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The datasets are sorted from earliest to latest release year. Writer Order GERMANA [122] 3.3.1 handwritten RODRIGO [135] 3.3.2 handwritten IAM-HistDB [47] 3.3.3 handwritten PHTD [2] 3.2.1 handwritten PBOK [5] 3.2.2 handwritten IMPACT [117] 3.2.3 both ESPOSALLES [128] 3.3.4 handwritten BH2M [44] 3.3.5 handwritten HADARA80P [116] 3.3.6 handwritten ENP [29] 3.2.4 printed GRPOLY-DB [55] 3.2.5 both DocExplore [41] 3.3.7 both DIVA-HisDB [144] 3.2.6 handwritten AMADI_LontarSet [77] 3.3.8 handwritten ICFHR16 CLaMM [33] 3.1.1 handwritten ICDAR17 CLaMM [34] 3.1.2 handwritten HBA 1.0 [101] 3.2.7 both SleukRith [153] 3.3.9 handwritten VML-HD [76] 3.3.10 handwritten CFRAMUZ [10] 3.3.11 handwritten Lontar Sunda [148] 3.3.12 handwritten ICDAR17 REID2017 [30] 3.3.13 printed ICDAR17 Historical-WI [45] 3.3.14 handwritten Kuzushiji [27] 3.3.15 printed READ-BAD [60] 3.2.8 printed Warped Arabic [40] 3.2.9 both MHDID [140] 3.3.16 handwritten ( ) Tripitaka [161] 3.3.17 handwritten KERTAS [1] 3.1.3 handwritten ICFHR18 RASM2018 [31] 3.3.18 handwritten ICFHR18 Asian Palm Leaf [78] 3.3.19 handwritten ARDIS [84] 3.3.20 handwritten Pinkas [83] 3.2.10 handwritten BADAM [79] 3.2.11 handwritten HORAE [14] 3.2.12 handwritten ICDAR19 cTDaR19 [52] 3.2.13 handwritten ICDAR19 DMAS2019 4 3.2.14 printed OBC306 [67] 3.3.21 han...…”

Section: Datasetsmentioning

confidence: 99%

See 1 more Smart Citation

A Survey of Historical Document Image Datasets

Nikolaidou¹,

Seuret²,

Mokayed³

et al. 2022

Preprint

View full text Add to dashboard Cite

This paper presents a systematic literature review of image datasets for document image analysis, focusing on historical documents, such as handwritten manuscripts and early prints. Finding appropriate datasets for historical document analysis is a crucial prerequisite to facilitate research using different machine learning algorithms. However, because of the very large variety of the actual data (e.g., scripts, tasks, dates, support systems, and amount of deterioration), the different formats for data and label representation, and the different evaluation processes and benchmarks, finding appropriate datasets is a difficult task. This work fills this gap, presenting a meta-study on existing datasets. After a systematic selection process (according to PRISMA guidelines), we select 56 studies that are chosen based on different factors, such as the year of publication, number of methods implemented in the article, reliability of the chosen algorithms, dataset size, and journal outlet. We summarize each study by assigning it to one of three pre-defined tasks: document classification, layout structure, or semantic analysis. We present the statistics, document type, language, tasks, input visual aspects, and ground truth information for every dataset. In addition, we provide the benchmark tasks and results from these papers or recent competitions. We further discuss gaps and challenges in this domain. We advocate for providing conversion tools to common formats (e.g., COCO format for computer vision tasks) and always providing a set of evaluation metrics, instead of just one, to make results comparable across studies.

show abstract

Section: Datasetsmentioning

confidence: 99%

“…OBC306 [67] is a dataset of 309,551 images for Oracle Bone character recognition distributed across 306 character classes. This dataset consists of patch samples derived from different sources from full image publications of oracle bones.…”

Section: Obc306mentioning

confidence: 99%

A Survey of Historical Document Image Datasets

Nikolaidou¹,

Seuret²,

Mokayed³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Recently, CNNs have achieved great progress in some computer vision tasks and are introduced into oracle character recognition. Huang et al [4] released a dataset of scanned oracle characters called OBC306 and presented the CNN-based evaluation for this dataset to serve as a benchmark. OracleNet [26] considered the radical-level composition of oracle characters, and detected radicals using a Capsule network [27] to recognize characters.…”

Section: A Oracle Character Recognitionmentioning

confidence: 99%

“…Research on oracle characters began in the late of 1890s shortly after oracle bones were unearthed. Thus far, nearly 4,500 different oracle characters have been discovered, and only about 2,200 characters have been successfully deciphered [3], [4]. To help with the excavation of new oracle bones and the identification of unseen characters, deep convolutional Mei Wang and Weihong Deng are with the Pattern Recognition and Intelligent System Laboratory, School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing, 100876, China.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Unsupervised Structure-Texture Separation Network for Oracle Character Recognition

Wang,

Deng,

Liu

2022

Preprint

View full text Add to dashboard Cite

Oracle bone script is the earliest-known Chinese writing system of the Shang dynasty and is precious to archeology and philology. However, real-world scanned oracle data are rare and few experts are available for annotation which make the automatic recognition of scanned oracle characters become a challenging task. Therefore, we aim to explore unsupervised domain adaptation to transfer knowledge from handprinted oracle data, which are easy to acquire, to scanned domain. We propose a structure-texture separation network (STSN), which is an end-to-end learning framework for joint disentanglement, transformation, adaptation and recognition. First, STSN disentangles features into structure (glyph) and texture (noise) components by generative models, and then aligns handprinted and scanned data in structure feature space such that the negative influence caused by serious noises can be avoided when adapting. Second, transformation is achieved via swapping the learned textures across domains and a classifier for final classification is trained to predict the labels of the transformed scanned characters. This not only guarantees the absolute separation, but also enhances the discriminative ability of the learned features. Extensive experiments on Oracle-241 dataset show that STSN outperforms other adaptation methods and successfully improves recognition performance on scanned data even when they are contaminated by long burial and careless excavation.

show abstract