{ "id": 122604, "title": "Bengali Data Sets & Publications! ", "url": "/competitions/bengaliai-cv19/discussion/122604", "author_name": "Innat", "post_date": "2019-12-21T12:11:25.336000", "votes": 35, "comment_count": 11, "views": 0, "content": "

Data Set ( Both Character and Numeral)

\n\n

Publication

\n

Year 2018

\n
    \n

  1. (Character) BornoNet: Bangla Handwritten Characters Recognition Using Convolutional Neural Network Article

    2. \n

  2. (Character + Digit) EkushNet: Using Convolutional Neural Network for Bangla Handwritten Recognition Article

    3. \n

  3. (Digit) Bangla Handwritten Digit Recognition Using Convolutional Neural Network Article

    4. \n

  4. (Digit) Recognition of Bengali Handwritten Digits Using Convolutional Neural Network Architectures. Article

    5. \n

  5. (Digit) Unconventional Wisdom: A New Transfer Learning Approach Applied to Bengali Numeral Classification Article

    6. \n

  6. (Digit) Bangla Handwritten Digit Recognition Approach with an Ensemble of Deep Residual Networks Article

    7. \n

  7. (Digit) Handwritten Bangla Numeral Recognition Using Ensembling of Convolutional Neural Network Article

    8. \n

  8. (Digit) Bangla Handwritten Digit Recognition Using Deep CNN for Large and Unbiased Dataset Article

    9. \n
\n

Year 2019

\n
    \n

  1. (Character) Basic to Compound: A Novel Transfer Learning Approach for Bengali Handwritten Character Recognition Article

    2. \n

  2. (Character) AIBangla: A Benchmark Dataset for Isolated Bangla Handwritten Basic and Compound Character Recognition Article

    3. \n

  3. (Digit) Recognizing Bangla Handwritten Numeral Utilizing Deep Long Short Term Memory Article

    4. \n

  4. (Digit) BDNet: Bengali handwritten numeral digit recognition based on densely connected convolutional neural networks Article

    5. \n

  5. (Character) Bangla Handwritten Character Recognition using Convolutional Neural Network with Data Augmentation Article

    6. \n
\n

Some of my Bengali brothers may find their own recent article in the above, awesome.😀
\nAnyway, Good luck everybody.

\n
\n

Update:

\n

You may find them useful,

\n

For Graphene dataset: https://www.kaggle.com/ipythonx/keras-grapheme-gridmask-augmix-ensemble

\n

For Digit dataset: https://www.kaggle.com/ipythonx/bd-numerals-digit-classification-in-capsule-nets

\n

Data Sets

\n

Numerals

\n\n

Character

\n", "messages": [ { "id": 700071, "postDate": "2019-12-21T12:11:25.337Z", "content": "

Data Set ( Both Character and Numeral)

\n\n

Publication

\n

Year 2018

\n
    \n

  1. (Character) BornoNet: Bangla Handwritten Characters Recognition Using Convolutional Neural Network Article

    2. \n

  2. (Character + Digit) EkushNet: Using Convolutional Neural Network for Bangla Handwritten Recognition Article

    3. \n

  3. (Digit) Bangla Handwritten Digit Recognition Using Convolutional Neural Network Article

    4. \n

  4. (Digit) Recognition of Bengali Handwritten Digits Using Convolutional Neural Network Architectures. Article

    5. \n

  5. (Digit) Unconventional Wisdom: A New Transfer Learning Approach Applied to Bengali Numeral Classification Article

    6. \n

  6. (Digit) Bangla Handwritten Digit Recognition Approach with an Ensemble of Deep Residual Networks Article

    7. \n

  7. (Digit) Handwritten Bangla Numeral Recognition Using Ensembling of Convolutional Neural Network Article

    8. \n

  8. (Digit) Bangla Handwritten Digit Recognition Using Deep CNN for Large and Unbiased Dataset Article

    9. \n
\n

Year 2019

\n
    \n

  1. (Character) Basic to Compound: A Novel Transfer Learning Approach for Bengali Handwritten Character Recognition Article

    2. \n

  2. (Character) AIBangla: A Benchmark Dataset for Isolated Bangla Handwritten Basic and Compound Character Recognition Article

    3. \n

  3. (Digit) Recognizing Bangla Handwritten Numeral Utilizing Deep Long Short Term Memory Article

    4. \n

  4. (Digit) BDNet: Bengali handwritten numeral digit recognition based on densely connected convolutional neural networks Article

    5. \n

  5. (Character) Bangla Handwritten Character Recognition using Convolutional Neural Network with Data Augmentation Article

    6. \n
\n

Some of my Bengali brothers may find their own recent article in the above, awesome.😀
\nAnyway, Good luck everybody.

\n
\n

Update:

\n

You may find them useful,

\n

For Graphene dataset: https://www.kaggle.com/ipythonx/keras-grapheme-gridmask-augmix-ensemble

\n

For Digit dataset: https://www.kaggle.com/ipythonx/bd-numerals-digit-classification-in-capsule-nets

\n

Data Sets

\n

Numerals

\n\n

Character

\n", "rawMarkdown": "## Data Set ( Both Character and Numeral)\n- [BanglaLekha-Isolated-Data-Set](https://data.mendeley.com/datasets/hf6sf8zrkc/2?__cf_chl_captcha_tk__=0eef3f155bb8cdbd4318462de9ce43b06e8b36fe-1576926645-0-ARwIJOlc0pqWADHpP7QJNSnlLmWPdU95TWcw8RABhCTk-MHuKadBfXUxQednw2omNlWNN2YFffaQjm6HFtzYhbKIaX6ZXIydnz6nMJUQL1p3MoCjLaD8C1Y7P2F33-MAWE3Eo1NstIjhrEzpMBoz4nmc0z8Fx1uftg5rDt5YCNCiWsBY4VcaijT_NXaReFueYu3CBgk-he-9c_Xr7FhsPoQCpg8ThfHCmELrGY8f5EKmi5eFhBEWP6-1JpByjl4Z4Qr5-3zWmjn0uBsoPF2DKD6SE10QnF1hcUP5rHbSfV4Iv8mLUYoTLLeArXclMmeNEQuWKo_AgS5fAzLFuFCf_DI) | [Article](https://arxiv.org/abs/1703.10661)\n\n- [Ekush-Data-Set](https://shahariarrabby.github.io/ekush/#overview) | [Article](https://www.researchgate.net/publication/334488314_Ekush_A_Multipurpose_and_Multitype_Comprehensive_Database_for_Online_Off-Line_Bangla_Handwritten_Characters)\n\n- [NumtaDB](https://www.kaggle.com/BengaliAI/numta) | [Article](https://bengali.ai/wp-content/uploads/datasets/assembled-bangla-handwritten.pdf)\n\n- [ISI Image Data](https://www.isical.ac.in/~ujjwal/download/BanglaNumeral.html) | [Article](https://www.researchgate.net/publication/228938909_A_Complete_Handwritten_Numeral_Database_of_Bangla_-_A_Major_Indic_Script)\n\n- [CMATERdb](https://code.google.com/archive/p/cmaterdb/)\n\n## Publication \n\nYear **2018**\n\n1. (Character) `BornoNet: Bangla Handwritten Characters Recognition Using Convolutional Neural Network` [Article](https://www.researchgate.net/publication/329048432_BornoNet_Bangla_Handwritten_Characters_Recognition_Using_Convolutional_Neural_Network)\n\n2. (Character + Digit) `EkushNet: Using Convolutional Neural Network for Bangla Handwritten Recognition` [Article](file:///D:/Research%20Zone/Bangla%20Digit%20Classifier/Related%20Paper/2018/EkushNet.pdf)\n\n3. (Digit) `Bangla Handwritten Digit Recognition Using Convolutional Neural Network` [Article](https://www.researchgate.net/publication/329606551_Bangla_Handwritten_Digit_Recognition_Using_Convolutional_Neural_Network) \n\n4. (Digit) `Recognition of Bengali Handwritten Digits Using Convolutional Neural Network Architectures.` [Article](https://ieeexplore.ieee.org/document/8554753) \n\n5. (Digit) `Unconventional Wisdom: A New Transfer Learning Approach Applied to Bengali Numeral Classification` [Article](https://www.researchgate.net/publication/326989744_Unconventional_Wisdom_A_New_Transfer_Learning_Approach_Applied_to_Bengali_Numeral_Classification)\n\n6. (Digit) `Bangla Handwritten Digit Recognition Approach with an Ensemble of Deep Residual Networks ` [Article](https://ieeexplore.ieee.org/abstract/document/8554674)\n\n7. (Digit) `Handwritten Bangla Numeral Recognition Using Ensembling of Convolutional Neural Network` [Article](https://ieeexplore.ieee.org/document/8631944)\n\n8. (Digit) `Bangla Handwritten Digit Recognition Using Deep CNN for Large and Unbiased Dataset` [Article](https://www.researchgate.net/publication/329391937_Bangla_Handwritten_Digit_Recognition_Using_Deep_CNN_for_Large_and_Unbiased_Dataset)\n\n#### Year **2019**\n\n1. (Character) `Basic to Compound: A Novel Transfer Learning Approach for Bengali Handwritten Character Recognition` [Article](https://www.researchgate.net/publication/336287494_Basic_to_Compound_A_Novel_Transfer_Learning_Approach_for_Bengali_Handwritten_Character_Recognition)\n\n2. (Character) `AIBangla: A Benchmark Dataset for Isolated Bangla Handwritten Basic and Compound Character Recognition` [Article](https://www.researchgate.net/publication/337499773_AIBangla_A_Benchmark_Dataset_for_Isolated_Bangla_Handwritten_Basic_and_Compound_Character_Recognition)\n\n3. (Digit) `Recognizing Bangla Handwritten Numeral Utilizing Deep Long Short Term Memory` [Article](http://www.mecs-press.org/ijigsp/ijigsp-v11-n1/IJIGSP-V11-N1-3.pdf)\n\n4. (Digit) `BDNet: Bengali handwritten numeral digit recognition based on densely connected convolutional neural networks` [Article](https://www.researchgate.net/publication/333679021_BDNet_Bengali_handwritten_numeral_digit_recognition_based_on_densely_connected_convolutional_neural_networks)\n\n5. (Character) `Bangla Handwritten Character Recognition using Convolutional Neural Network with Data Augmentation` [Article](https://www.researchgate.net/publication/333618625_Bangla_Handwritten_Character_Recognition_using_Convolutional_Neural_Network_with_Data_Augmentation)\n\nSome of my Bengali brothers may find their own recent article in the above, awesome.😀\nAnyway, Good luck everybody. \n\n---\n\n## Update:\nYou may find them useful,\n\nFor Graphene dataset: https://www.kaggle.com/ipythonx/keras-grapheme-gridmask-augmix-ensemble\n\nFor Digit dataset: https://www.kaggle.com/ipythonx/bd-numerals-digit-classification-in-capsule-nets\n\n## Data Sets\n\n**Numerals**\n- [Ekush: Bangla Handwritten Data - Numerals](https://www.kaggle.com/ipythonx/ekush-bangla-handwritten-data-numerals)\n- [BanglaLekha-Isolated-Numerals](https://www.kaggle.com/ipythonx/banglalekhaisolatednumerals)\n- [NumtaDB: Bengali Handwritten Digits](https://www.kaggle.com/BengaliAI/numta)\n\n**Character**\n- [Bangla Grapheme Image 137x236](https://www.kaggle.com/ipythonx/137x236)\n", "votes": 35 }, { "id": 809005, "postDate": "2020-04-15T19:04:54.327Z", "content": "

Hello .Is there any progress on fake news detection from bangla dataset !

", "rawMarkdown": "Hello .Is there any progress on fake news detection from bangla dataset !", "votes": 1, "replies": [ { "id": 809525, "postDate": "2020-04-16T09:09:08.933Z", "content": "

@fahmidalizapiya sorry, I didn't know any well-known data set regarding this. However, here is a publication, you may ask the author if they are willing to collaborate. Along with this, check out this repo's data set section, it may come helpful to you.

", "rawMarkdown": "@fahmidalizapiya sorry, I didn't know any well-known data set regarding this. However, here is a [publication](http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/3619), you may ask the author if they are willing to collaborate. Along with this, check out [this repo's data set section](https://github.com/innat/ML-Bookmarks), it may come helpful to you." }, { "id": 809903, "postDate": "2020-04-16T15:10:25.933Z", "content": "

Yes there is no public dataset regarding this topic ! Thanks for the help though!

", "rawMarkdown": "Yes there is no public dataset regarding this topic ! Thanks for the help though!" } ] }, { "id": 737107, "postDate": "2020-02-04T23:29:50.490Z", "content": "

Thanks a lot for sharing @ipythonx this is very helpful.

", "rawMarkdown": "Thanks a lot for sharing @ipythonx this is very helpful.", "votes": 1 }, { "id": 706939, "postDate": "2019-12-31T03:35:24.467Z", "content": "

another alternative. since the data collection process is open sourced, you can collect your own data!\nhttps://github.com/BengaliAI/graphemePrepare

", "rawMarkdown": "another alternative. since the data collection process is open sourced, you can collect your own data!\nhttps://github.com/BengaliAI/graphemePrepare", "votes": 1 }, { "id": 704649, "postDate": "2019-12-27T19:25:25.053Z", "content": "

U can use open source bengali fonts as train data too

", "rawMarkdown": "U can use open source bengali fonts as train data too", "votes": 1, "replies": [ { "id": 704652, "postDate": "2019-12-27T19:33:20.040Z", "rawMarkdown": "", "isDeleted": true } ] }, { "id": 700420, "postDate": "2019-12-22T01:37:51.843Z", "content": "

Thanks I might try these models for fun :-) I will cite your post if it will

", "rawMarkdown": "Thanks I might try these models for fun :-) I will cite your post if it will", "votes": 2, "replies": [ { "id": 700783, "postDate": "2019-12-22T15:36:29.723Z", "content": "

I would recommend to check you to the model EkushNet (2018) and some published work in 2019. EkushNet was designed for both digit and character and (probably) some compound character and some modifiers.

", "rawMarkdown": "I would recommend to check you to the model EkushNet (2018) and some published work in 2019. EkushNet was designed for both digit and character and (probably) some compound character and some modifiers. ", "votes": 1 }, { "id": 702479, "postDate": "2019-12-24T18:25:04.190Z", "content": "

I tried EkushNet customized a bit(structure, activation func etc...) performance was um... 90~95%

", "rawMarkdown": "I tried EkushNet customized a bit(structure, activation func etc...) performance was um... 90~95%", "votes": 2 } ] }, { "id": 709773, "postDate": "2020-01-03T23:02:16.040Z", "content": "

Thanks for sharing!

", "rawMarkdown": "Thanks for sharing!", "votes": 1 } ], "comments": [ { "id": 809005, "author_name": "FAHMIDA LIZA PIYA", "author_url": "", "post_date": "2020-04-15T19:04:54.327000", "content": "

Hello .Is there any progress on fake news detection from bangla dataset !

", "votes": 1, "replies": [ { "id": 809525, "author_name": "Innat", "author_url": "", "post_date": "2020-04-16T09:09:08.933000", "content": "

@fahmidalizapiya sorry, I didn't know any well-known data set regarding this. However, here is a publication, you may ask the author if they are willing to collaborate. Along with this, check out this repo's data set section, it may come helpful to you.

", "votes": 0, "replies": [] }, { "id": 809903, "author_name": "FAHMIDA LIZA PIYA", "author_url": "", "post_date": "2020-04-16T15:10:25.933000", "content": "

Yes there is no public dataset regarding this topic ! Thanks for the help though!

", "votes": 0, "replies": [] } ] }, { "id": 737107, "author_name": "Rohit Agarwal", "author_url": "", "post_date": "2020-02-04T23:29:50.490000", "content": "

Thanks a lot for sharing @ipythonx this is very helpful.

", "votes": 1, "replies": [] }, { "id": 706939, "author_name": "hengck23", "author_url": "", "post_date": "2019-12-31T03:35:24.467000", "content": "

another alternative. since the data collection process is open sourced, you can collect your own data!\nhttps://github.com/BengaliAI/graphemePrepare

", "votes": 1, "replies": [] }, { "id": 704649, "author_name": "hengck23", "author_url": "", "post_date": "2019-12-27T19:25:25.053000", "content": "

U can use open source bengali fonts as train data too

", "votes": 1, "replies": [ { "id": 704652, "author_name": "", "author_url": "", "post_date": "2019-12-27T19:33:20.040000", "content": "", "votes": 0, "replies": [] } ] }, { "id": 700420, "author_name": "Hanjoon Choe", "author_url": "", "post_date": "2019-12-22T01:37:51.843000", "content": "

Thanks I might try these models for fun :-) I will cite your post if it will

", "votes": 2, "replies": [ { "id": 700783, "author_name": "Innat", "author_url": "", "post_date": "2019-12-22T15:36:29.723000", "content": "

I would recommend to check you to the model EkushNet (2018) and some published work in 2019. EkushNet was designed for both digit and character and (probably) some compound character and some modifiers.

", "votes": 1, "replies": [] }, { "id": 702479, "author_name": "Hanjoon Choe", "author_url": "", "post_date": "2019-12-24T18:25:04.190000", "content": "

I tried EkushNet customized a bit(structure, activation func etc...) performance was um... 90~95%

", "votes": 2, "replies": [] } ] }, { "id": 709773, "author_name": "PY", "author_url": "", "post_date": "2020-01-03T23:02:16.040000", "content": "

Thanks for sharing!

", "votes": 1, "replies": [] } ], "raw_markdown_by_id": { "700071": "## Data Set ( Both Character and Numeral)\n- [BanglaLekha-Isolated-Data-Set](https://data.mendeley.com/datasets/hf6sf8zrkc/2?__cf_chl_captcha_tk__=0eef3f155bb8cdbd4318462de9ce43b06e8b36fe-1576926645-0-ARwIJOlc0pqWADHpP7QJNSnlLmWPdU95TWcw8RABhCTk-MHuKadBfXUxQednw2omNlWNN2YFffaQjm6HFtzYhbKIaX6ZXIydnz6nMJUQL1p3MoCjLaD8C1Y7P2F33-MAWE3Eo1NstIjhrEzpMBoz4nmc0z8Fx1uftg5rDt5YCNCiWsBY4VcaijT_NXaReFueYu3CBgk-he-9c_Xr7FhsPoQCpg8ThfHCmELrGY8f5EKmi5eFhBEWP6-1JpByjl4Z4Qr5-3zWmjn0uBsoPF2DKD6SE10QnF1hcUP5rHbSfV4Iv8mLUYoTLLeArXclMmeNEQuWKo_AgS5fAzLFuFCf_DI) | [Article](https://arxiv.org/abs/1703.10661)\n\n- [Ekush-Data-Set](https://shahariarrabby.github.io/ekush/#overview) | [Article](https://www.researchgate.net/publication/334488314_Ekush_A_Multipurpose_and_Multitype_Comprehensive_Database_for_Online_Off-Line_Bangla_Handwritten_Characters)\n\n- [NumtaDB](https://www.kaggle.com/BengaliAI/numta) | [Article](https://bengali.ai/wp-content/uploads/datasets/assembled-bangla-handwritten.pdf)\n\n- [ISI Image Data](https://www.isical.ac.in/~ujjwal/download/BanglaNumeral.html) | [Article](https://www.researchgate.net/publication/228938909_A_Complete_Handwritten_Numeral_Database_of_Bangla_-_A_Major_Indic_Script)\n\n- [CMATERdb](https://code.google.com/archive/p/cmaterdb/)\n\n## Publication \n\nYear **2018**\n\n1. (Character) `BornoNet: Bangla Handwritten Characters Recognition Using Convolutional Neural Network` [Article](https://www.researchgate.net/publication/329048432_BornoNet_Bangla_Handwritten_Characters_Recognition_Using_Convolutional_Neural_Network)\n\n2. (Character + Digit) `EkushNet: Using Convolutional Neural Network for Bangla Handwritten Recognition` [Article](file:///D:/Research%20Zone/Bangla%20Digit%20Classifier/Related%20Paper/2018/EkushNet.pdf)\n\n3. (Digit) `Bangla Handwritten Digit Recognition Using Convolutional Neural Network` [Article](https://www.researchgate.net/publication/329606551_Bangla_Handwritten_Digit_Recognition_Using_Convolutional_Neural_Network) \n\n4. (Digit) `Recognition of Bengali Handwritten Digits Using Convolutional Neural Network Architectures.` [Article](https://ieeexplore.ieee.org/document/8554753) \n\n5. (Digit) `Unconventional Wisdom: A New Transfer Learning Approach Applied to Bengali Numeral Classification` [Article](https://www.researchgate.net/publication/326989744_Unconventional_Wisdom_A_New_Transfer_Learning_Approach_Applied_to_Bengali_Numeral_Classification)\n\n6. (Digit) `Bangla Handwritten Digit Recognition Approach with an Ensemble of Deep Residual Networks ` [Article](https://ieeexplore.ieee.org/abstract/document/8554674)\n\n7. (Digit) `Handwritten Bangla Numeral Recognition Using Ensembling of Convolutional Neural Network` [Article](https://ieeexplore.ieee.org/document/8631944)\n\n8. (Digit) `Bangla Handwritten Digit Recognition Using Deep CNN for Large and Unbiased Dataset` [Article](https://www.researchgate.net/publication/329391937_Bangla_Handwritten_Digit_Recognition_Using_Deep_CNN_for_Large_and_Unbiased_Dataset)\n\n#### Year **2019**\n\n1. (Character) `Basic to Compound: A Novel Transfer Learning Approach for Bengali Handwritten Character Recognition` [Article](https://www.researchgate.net/publication/336287494_Basic_to_Compound_A_Novel_Transfer_Learning_Approach_for_Bengali_Handwritten_Character_Recognition)\n\n2. (Character) `AIBangla: A Benchmark Dataset for Isolated Bangla Handwritten Basic and Compound Character Recognition` [Article](https://www.researchgate.net/publication/337499773_AIBangla_A_Benchmark_Dataset_for_Isolated_Bangla_Handwritten_Basic_and_Compound_Character_Recognition)\n\n3. (Digit) `Recognizing Bangla Handwritten Numeral Utilizing Deep Long Short Term Memory` [Article](http://www.mecs-press.org/ijigsp/ijigsp-v11-n1/IJIGSP-V11-N1-3.pdf)\n\n4. (Digit) `BDNet: Bengali handwritten numeral digit recognition based on densely connected convolutional neural networks` [Article](https://www.researchgate.net/publication/333679021_BDNet_Bengali_handwritten_numeral_digit_recognition_based_on_densely_connected_convolutional_neural_networks)\n\n5. (Character) `Bangla Handwritten Character Recognition using Convolutional Neural Network with Data Augmentation` [Article](https://www.researchgate.net/publication/333618625_Bangla_Handwritten_Character_Recognition_using_Convolutional_Neural_Network_with_Data_Augmentation)\n\nSome of my Bengali brothers may find their own recent article in the above, awesome.😀\nAnyway, Good luck everybody. \n\n---\n\n## Update:\nYou may find them useful,\n\nFor Graphene dataset: https://www.kaggle.com/ipythonx/keras-grapheme-gridmask-augmix-ensemble\n\nFor Digit dataset: https://www.kaggle.com/ipythonx/bd-numerals-digit-classification-in-capsule-nets\n\n## Data Sets\n\n**Numerals**\n- [Ekush: Bangla Handwritten Data - Numerals](https://www.kaggle.com/ipythonx/ekush-bangla-handwritten-data-numerals)\n- [BanglaLekha-Isolated-Numerals](https://www.kaggle.com/ipythonx/banglalekhaisolatednumerals)\n- [NumtaDB: Bengali Handwritten Digits](https://www.kaggle.com/BengaliAI/numta)\n\n**Character**\n- [Bangla Grapheme Image 137x236](https://www.kaggle.com/ipythonx/137x236)\n", "809005": "Hello .Is there any progress on fake news detection from bangla dataset !", "737107": "Thanks a lot for sharing @ipythonx this is very helpful.", "706939": "another alternative. since the data collection process is open sourced, you can collect your own data!\nhttps://github.com/BengaliAI/graphemePrepare", "704649": "U can use open source bengali fonts as train data too", "700420": "Thanks I might try these models for fun :-) I will cite your post if it will", "709773": "Thanks for sharing!" } }