Sökning: id:"swepub:oai:DiVA.org:bth-17741" >
ARDIS :
-
Kusetogullari, Hüseyin,1981-Blekinge Tekniska Högskola,Institutionen för datalogi och datorsystemteknik,Big Data
(författare)
ARDIS : A Swedish Historical Handwritten Digit Dataset
- Artikel/kapitelEngelska2020
Förlag, utgivningsår, omfång ...
-
2019-03-29
-
Springer Nature Switzerland,2020
-
electronicrdacarrier
Nummerbeteckningar
-
LIBRIS-ID:oai:DiVA.org:bth-17741
-
https://urn.kb.se/resolve?urn=urn:nbn:se:bth-17741URI
-
https://doi.org/10.1007/s00521-019-04163-3DOI
Kompletterande språkuppgifter
-
Språk:engelska
-
Sammanfattning på:engelska
Ingår i deldatabas
Klassifikation
-
Ämneskategori:ref swepub-contenttype
-
Ämneskategori:art swepub-publicationtype
Anmärkningar
-
open access
-
This paper introduces a new image-based handwrittenhistorical digit dataset named ARDIS (Arkiv DigitalSweden). The images in ARDIS dataset are extractedfrom 15,000 Swedish church records which were writtenby different priests with various handwriting styles in thenineteenth and twentieth centuries. The constructed datasetconsists of three single digit datasets and one digit stringsdataset. The digit strings dataset includes 10,000 samplesin Red-Green-Blue (RGB) color space, whereas, the otherdatasets contain 7,600 single digit images in different colorspaces. An extensive analysis of machine learning methodson several digit datasets is examined. Additionally, correlationbetween ARDIS and existing digit datasets ModifiedNational Institute of Standards and Technology (MNIST)and United States Postal Service (USPS) is investigated. Experimental results show that machine learning algorithms,including deep learning methods, provide low recognitionaccuracy as they face difficulties when trained on existingdatasets and tested on ARDIS dataset. Accordingly, ConvolutionalNeural Network (CNN) trained on MNIST andUSPS and tested on ARDIS provide the highest accuracies 58.80% and 35.44%, respectively. Consequently, the resultsreveal that machine learning methods trained on existingdatasets can have difficulties to recognize digits effectivelyon our dataset which proves that ARDIS dataset hasunique characteristics. This dataset is publicly available forthe research community to further advance handwritten digitrecognition algorithms.
Ämnesord och genrebeteckningar
Biuppslag (personer, institutioner, konferenser, titlar ...)
-
Yavariabdi, AmirKTO Karatay University, TUR,Department of Mechatronics Engineering
(författare)
-
Cheddad, AbbasBlekinge Tekniska Högskola,Institutionen för datalogi och datorsystemteknik,Big Data(Swepub:bth)abc
(författare)
-
Grahn, HåkanBlekinge Tekniska Högskola,Institutionen för datalogi och datorsystemteknik(Swepub:bth)hgr
(författare)
-
Johan, HallArkiv Digital, SWE
(författare)
-
Blekinge Tekniska HögskolaInstitutionen för datalogi och datorsystemteknik
(creator_code:org_t)
Sammanhörande titlar
-
Ingår i:Neural Computing & Applications: Springer Nature Switzerland32:21, s. 16505-165180941-06431433-3058
Internetlänk
Hitta via bibliotek
Till lärosätets databas