![\"\"](\"https://www.researchgate.net/profile/Nibaran_Das/publication/220888578/figure/fig1/AS:305493600948232@1449846758358/Some-obsolete-characters-which-are-not-used-in-current-Bangla-Text.png\")
my idea is to do something like this:
\n\n
i find the Ekush Dataset, which is hand written alphabets.\nhttps://shahariarrabby.github.io/ekush\nhttps://www.ukessays.com/dissertation/full-dissertations/bangla-handwritten-character-data-repository.php
\n\n![\"\"](\"https://images.ukdissertations.com/33/0454117.013.jpg\")
another dataset: \nBanglaLekha-Isolated: A Comprehensive Bangla Handwritten Character Dataset\nhttps://arxiv.org/abs/1703.10661
\n\nBeing unfamiliar with bengali language, I would like to know if my approach is correct or feasible. Thanks.
", "messages": [ { "id": "706209", "postDate": "12/30/2019 04:31:08", "content": "my idea is to do something like this:
\n\n
i find the Ekush Dataset, which is hand written alphabets.\nhttps://shahariarrabby.github.io/ekush\nhttps://www.ukessays.com/dissertation/full-dissertations/bangla-handwritten-character-data-repository.php
\n\n
another dataset: \nBanglaLekha-Isolated: A Comprehensive Bangla Handwritten Character Dataset\nhttps://arxiv.org/abs/1703.10661
\n\nBeing unfamiliar with bengali language, I would like to know if my approach is correct or feasible. Thanks.
", "rawMarkdown": "my idea is to do something like this:\n\n\n\n\ni find the Ekush Dataset, which is hand written alphabets.\nhttps://shahariarrabby.github.io/ekush\nhttps://www.ukessays.com/dissertation/full-dissertations/bangla-handwritten-character-data-repository.php\n\n\n\n\nanother dataset: \nBanglaLekha-Isolated: A Comprehensive Bangla Handwritten Character Dataset\nhttps://arxiv.org/abs/1703.10661\n\nBeing unfamiliar with bengali language, I would like to know if my approach is correct or feasible. Thanks.", "votes": null }, { "id": "706211", "postDate": "12/30/2019 04:35:09", "content": "I also interested in this idea. Need help from whom knows this language.
", "rawMarkdown": "I also interested in this idea. Need help from whom knows this language.", "votes": null }, { "id": "706218", "postDate": "12/30/2019 04:51:31", "content": "Lesson 1 | জ | ন | ল | ফ | র | া | ে | ি| Alphabet | Learn Bengali with Baneebee | Shusmita Shyama\nhttps://www.youtube.com/watch?v=S-BTrTF8BoY
", "rawMarkdown": "Lesson 1 | জ | ন | ল | ফ | র | া | ে | ি| Alphabet | Learn Bengali with Baneebee | Shusmita Shyama\nhttps://www.youtube.com/watch?v=S-BTrTF8BoY", "votes": null }, { "id": "707684", "postDate": "01/01/2020 09:43:54", "content": "Are you thinking about synthesizing new graphemes using the characters from these datasets? Or are you thinking about using some of the isolated characters from these datasets as possible cases of graphemes with no diacritics (i.e. corresponding grapheme root with 0 vowel and 0 consonant diacritic labels)?
", "rawMarkdown": "Are you thinking about synthesizing new graphemes using the characters from these datasets? Or are you thinking about using some of the isolated characters from these datasets as possible cases of graphemes with no diacritics (i.e. corresponding grapheme root with 0 vowel and 0 consonant diacritic labels)?", "votes": null }, { "id": "709835", "postDate": "01/04/2020 01:31:20", "content": "I am interested in this idea and the project and also it's my native language, however, I am a beginner in deep learning and kaggle, but would kindly like to offer my help if I can anyhow help. Thanks
", "rawMarkdown": "I am interested in this idea and the project and also it's my native language, however, I am a beginner in deep learning and kaggle, but would kindly like to offer my help if I can anyhow help. Thanks", "votes": null }, { "id": "710032", "postDate": "01/04/2020 08:39:16", "content": "Nice
", "rawMarkdown": "Nice", "votes": null }, { "id": "710874", "postDate": "01/05/2020 11:20:16", "content": "The ekush dataset has both male and female handwriting an it has mostly grapheme root and diacritics but not the combinations of them I believe . \nI am not quite clear as to what you are proposing .
\n\nif you are trying to find the distance between two graphemes by manually constructing the graphemes from root ?
\n\nare you trying to add more data to the existing dataset and find more training data and grapheme combination ?
\n\nby the way , In your example the first one is a Grapheme consisting of Grapheme Root + Diacritics but others are grapheme root itself .
", "rawMarkdown": "The ekush dataset has both male and female handwriting an it has mostly grapheme root and diacritics but not the combinations of them I believe . \nI am not quite clear as to what you are proposing . \n\nif you are trying to find the distance between two graphemes by manually constructing the graphemes from root ?\n\nare you trying to add more data to the existing dataset and find more training data and grapheme combination ?\n\nby the way , In your example the first one is a Grapheme consisting of Grapheme Root + Diacritics but others are grapheme root itself .", "votes": null }, { "id": "711463", "postDate": "01/06/2020 06:18:33", "content": "looks more complicated lol but its ok!
", "rawMarkdown": "looks more complicated lol but its ok!", "votes": null } ], "comments": [ { "id": 706211, "author_name": "ildoonet", "author_url": "", "post_date": "12/30/2019 04:35:09", "content": "I also interested in this idea. Need help from whom knows this language.
", "votes": null, "replies": [ { "id": 709835, "author_name": "hkarim", "author_url": "", "post_date": "01/04/2020 01:31:20", "content": "I am interested in this idea and the project and also it's my native language, however, I am a beginner in deep learning and kaggle, but would kindly like to offer my help if I can anyhow help. Thanks
", "votes": null, "replies": [] } ] }, { "id": 706218, "author_name": "hengck23", "author_url": "", "post_date": "12/30/2019 04:51:31", "content": "Lesson 1 | জ | ন | ল | ফ | র | া | ে | ি| Alphabet | Learn Bengali with Baneebee | Shusmita Shyama\nhttps://www.youtube.com/watch?v=S-BTrTF8BoY
", "votes": null, "replies": [] }, { "id": 707684, "author_name": "imtiazprio", "author_url": "", "post_date": "01/01/2020 09:43:54", "content": "Are you thinking about synthesizing new graphemes using the characters from these datasets? Or are you thinking about using some of the isolated characters from these datasets as possible cases of graphemes with no diacritics (i.e. corresponding grapheme root with 0 vowel and 0 consonant diacritic labels)?
", "votes": null, "replies": [] }, { "id": 710032, "author_name": "jchoong", "author_url": "", "post_date": "01/04/2020 08:39:16", "content": "Nice
", "votes": null, "replies": [] }, { "id": 710874, "author_name": "phoenix9032", "author_url": "", "post_date": "01/05/2020 11:20:16", "content": "The ekush dataset has both male and female handwriting an it has mostly grapheme root and diacritics but not the combinations of them I believe . \nI am not quite clear as to what you are proposing .
\n\nif you are trying to find the distance between two graphemes by manually constructing the graphemes from root ?
\n\nare you trying to add more data to the existing dataset and find more training data and grapheme combination ?
\n\nby the way , In your example the first one is a Grapheme consisting of Grapheme Root + Diacritics but others are grapheme root itself .
", "votes": null, "replies": [] }, { "id": 711463, "author_name": "seungjo2", "author_url": "", "post_date": "01/06/2020 06:18:33", "content": "looks more complicated lol but its ok!
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "706209": "my idea is to do something like this:\n\n\n\n\ni find the Ekush Dataset, which is hand written alphabets.\nhttps://shahariarrabby.github.io/ekush\nhttps://www.ukessays.com/dissertation/full-dissertations/bangla-handwritten-character-data-repository.php\n\n\n\n\nanother dataset: \nBanglaLekha-Isolated: A Comprehensive Bangla Handwritten Character Dataset\nhttps://arxiv.org/abs/1703.10661\n\nBeing unfamiliar with bengali language, I would like to know if my approach is correct or feasible. Thanks.", "706211": "I also interested in this idea. Need help from whom knows this language.", "706218": "Lesson 1 | জ | ন | ল | ফ | র | া | ে | ি| Alphabet | Learn Bengali with Baneebee | Shusmita Shyama\nhttps://www.youtube.com/watch?v=S-BTrTF8BoY", "707684": "Are you thinking about synthesizing new graphemes using the characters from these datasets? Or are you thinking about using some of the isolated characters from these datasets as possible cases of graphemes with no diacritics (i.e. corresponding grapheme root with 0 vowel and 0 consonant diacritic labels)?", "709835": "I am interested in this idea and the project and also it's my native language, however, I am a beginner in deep learning and kaggle, but would kindly like to offer my help if I can anyhow help. Thanks", "710032": "Nice", "710874": "The ekush dataset has both male and female handwriting an it has mostly grapheme root and diacritics but not the combinations of them I believe . \nI am not quite clear as to what you are proposing . \n\nif you are trying to find the distance between two graphemes by manually constructing the graphemes from root ?\n\nare you trying to add more data to the existing dataset and find more training data and grapheme combination ?\n\nby the way , In your example the first one is a Grapheme consisting of Grapheme Root + Diacritics but others are grapheme root itself .", "711463": "looks more complicated lol but its ok!" }, "source": "meta" }