Important dates:
- Feb 1, 2017 : Website online
- Feb 1, 2017 : Registration Open
- Feb 20, 2017: Sample dataset available
March 15, March 24, April 14
April 21, 2017: Training dataset availableMarch 25, April 3, April 21
April 28 2017: Validation dataset availableMarch 31, April 15, May 1
June 30, 2017: Registration closesApril 5, April 18, May 5,
June 30, 2017: Submission of systemsApril 7, April 20, May 10,
May 20, 2017: Test dataset availableApril 10, April 24, May 17,
June 30, 2017: Submission of results
Latest News
- Feb 1, 2017 : Website online, registration open
- Feb 20, 2017 : Sample Dataset available
- March 15, 2017 : Deadline extension
- March 31, 2017 : Following the extension of the ICDAR2017 full paper submission deadline, the training dataset will now be available on 14th April 2017 to the registered participants
- April 21, 2017 : The training dataset is now available to the registered participants
- April 30, 2017 : The validation dataset is now available to the registered participants
- May 29, 2017 : The test dataset is available after system submission
- May 29, 2017 : Competition deadline extend
Evaluation
Participants in any one or multiple task are welcome. Performance evaluation and system ranking will be based on single task. Each task will be evaluated and ranked separately. The winner will be ranked based on the identification accuracies. The system which performs best for individual tasks will be declared the winner. The test dataset is a closed dataset and will be made available once the participants submit their systems for evaluation.
Accuracy will be calculated as follows,
Accuracy = (CC/GT) X 100
Where, CC is the number of correctly identified samples; GT is the ground truth or total number of samples.
The participants have to submit their systems / executable either in Windows (XP or Win 7) executable format or Linux/Unix. The command lines of the system accepted for the competition are as follows,
1. Task 1 (Tri-Scripts identification with Indian scripts):
ClassifyEngHin<3rdScriptsName> Task1TestImages.txt Task1Result.txt
Where, is one of the acronym from {Ben, Ori, Guj, Pun, Kan, Tam, Tel, Arb, Mam}
Task1TestImages.txt is the input file having a list of paths to each JPG image to be tested:
Example:
D:\CVSI2017\ Task1TestImages \Test1.jpg
D:\CVSI2017\ Task1TestImages \Test2.jpg
Task1Result.txt is a single file which should have the following format and saves the identification result for all
the samples in the test dataset:
<testSample name> | <Identified script>
Example:
D:\CVSI2017\Task1TestImages\Test1.jpg |Eng
D:\CVSI2017\Task1TestImages\Test2.jpg |Ben
<Identified script> should be any one of the three character string from the set {Eng, Hin, Ben, Ori, Guj, Pun,
Kan, Tam, Tel, Arb, Mam, Chi, Tha, Jap, Kor}
2. Task 2 (Tri-Scripts identification with Oriental scripts)
ClassifyEng<2ndScripName><3rdScriptsName> Task2TestImages.txt Task2Result.txt
Where,<2ndScriptName> and <3rdScriptName> is one of the acronym from {Chi, Tha, Jap, Kor}
3. Task 3 (North Indian Scripts identification)
ClassifyNorthIndianScripts Task3TestImages.txt Task3Result.txt
4. Task 4 (South Indian Scripts identification)
ClassifySouthIndianScripts Task4TestImages.txt Task4Result.txt
5. Task 5 (South Asian Scripts identification)
ClassifyOrientalScripts Task5TestImages.txt Task5Result.txt
6. Task 6 (All Scripts identification)
ClassifyAllScripts Task6TestImages.txt Task6Result.txt
All the input text files and the result files have the same format as mentioned in point 1, above.