how to make inferences - Githubissues

songyh10 commented 4 years ago

By running the demo, we can get a visualization image with bounding boxes and characters. However, is there any inference command which can return the words or phrases instead of only characters? Thanks! @Real-YeJ

Real-YeJ commented 4 years ago

@songyh10 we save the contour points of word-level text as txt file. It‘s easy to draw the contours on images by python.

songyh10 commented 4 years ago

Then how do you compare the performance of TextFuseNet with other methods on the benchmark datasets, e.g. ICDAR2015 or Total-Text? Do you have some existing relevant inference code or tools to make this performance comparison? Thanks! @Real-YeJ

Real-YeJ commented 4 years ago

@songyh10 ic15: https://rrc.cvc.uab.es/?ch=4 totaltext: https://github.com/cs-chan/Total-Text-Dataset

animesh-007 commented 3 years ago

Then how do you compare the performance of TextFuseNet with other methods on the benchmark datasets, e.g. ICDAR2015 or Total-Text? Do you have some existing relevant inference code or tools to make this performance comparison? Thanks! @Real-YeJ

Hello @songyh10. Were you able to calculate F-measure, Recall, and Precision using this code. If so, Can you guide me on how to do the same?

Real-YeJ commented 3 years ago

@animesh-007 please pay attention to our update for evaluation

StudentChen commented 3 years ago

Then how do you compare the performance of TextFuseNet with other methods on the benchmark datasets, e.g. ICDAR2015 or Total-Text? Do you have some existing relevant inference code or tools to make this performance comparison? Thanks! @Real-YeJ

Hello @songyh10. Were you able to calculate F-measure, Recall, and Precision using this code. If so, Can you guide me on how to do the same?

hi,do you solve this problem? can you help me how to do the same? Thank you so much.

animesh-007 commented 3 years ago

Then how do you compare the performance of TextFuseNet with other methods on the benchmark datasets, e.g. ICDAR2015 or Total-Text? Do you have some existing relevant inference code or tools to make this performance comparison? Thanks! @Real-YeJ

Hello @songyh10. Were you able to calculate F-measure, Recall, and Precision using this code. If so, Can you guide me on how to do the same?

hi,do you solve this problem? can you help me how to do the same? Thank you so much.

Hi @StudentChen. You can calculate these metrics by visiting the links provided in the evaluation in the readme. After visiting those links you can find that they have provided a script for calculating these metrics.

StudentChen commented 3 years ago

Thank you you you,I will have a try according to your suggestion. So kind of you.

Then how do you compare the performance of TextFuseNet with other methods on the benchmark datasets, e.g. ICDAR2015 or Total-Text? Do you have some existing relevant inference code or tools to make this performance comparison? Thanks! @Real-YeJ

Hello @songyh10. Were you able to calculate F-measure, Recall, and Precision using this code. If so, Can you guide me on how to do the same?

hi,do you solve this problem? can you help me how to do the same? Thank you so much.

Hi @StudentChen. You can calculate these metrics by visiting the links provided in the evaluation in the readme. After visiting those links you can find that they have provided a script for calculating these metrics.

Thank you you you,I will have a try according to your suggestion. So kind of you.

animesh-007 commented 3 years ago

No problem. Let me know if you face any other issue. I will be happy to help.😇

StudentChen commented 3 years ago

hi,do you train this model by yourself? it needs the train dataset like the json file. do you get the train.json? Thank you for your answer.

------------------ 原始邮件 ------------------ 发件人: "ying09/TextFuseNet" <notifications@github.com>; 发送时间: 2021年1月7日(星期四) 中午11:04 收件人: "ying09/TextFuseNet"<TextFuseNet@noreply.github.com>; 抄送: "1571786636"<1571786636@qq.com>;"Mention"<mention@noreply.github.com>; 主题: Re: [ying09/TextFuseNet] how to make inferences (#18)

No problem. Let me know if you face any other issue. I will be happy to help.😇

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

animesh-007 commented 3 years ago

hi,do you train this model by yourself? it needs the train dataset like the json file. do you get the train.json? Thank you for your answer. … ------------------ 原始邮件 ------------------ 发件人: "ying09/TextFuseNet" <notifications@github.com>; 发送时间: 2021年1月7日(星期四) 中午11:04 收件人: "ying09/TextFuseNet"<TextFuseNet@noreply.github.com>; 抄送: "1571786636"<1571786636@qq.com>;"Mention"<mention@noreply.github.com>; 主题: Re: [ying09/TextFuseNet] how to make inferences (#18) No problem. Let me know if you face any other issue. I will be happy to help.😇 — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Hi @StudentChen, yes you have to create json files in coco format.

StudentChen commented 3 years ago

thank you，by the way，Could you please tell me how to make it? is there direct conversion tool or code? and the dataset format must like the author's annotation example? I have learned that there are some datasets' coco format json file online,but I am not sure it is suitable for this model. if it is convenient for you to send me your train.json through e-mail individually, I just want to study this model.Thank you so much.Anyway Good luck to you, thank you sincerely.

------------------ 原始邮件 ------------------ 发件人: "Animesh Gupta"<notifications@github.com>; 发送时间: 2021年2月6日(星期六) 晚上11:30 收件人: "ying09/TextFuseNet"<TextFuseNet@noreply.github.com>; 抄送: "1571786636"<1571786636@qq.com>; "Mention"<mention@noreply.github.com>; 主题: Re: [ying09/TextFuseNet] how to make inferences (#18)

hi,do you train this model by yourself? it needs the train dataset like the json file. do you get the train.json? Thank you for your answer. … ------------------ 原始邮件 ------------------ 发件人: "ying09/TextFuseNet" <notifications@github.com>; 发送时间: 2021年1月7日(星期四) 中午11:04 收件人: "ying09/TextFuseNet"<TextFuseNet@noreply.github.com>; 抄送: "1571786636"<1571786636@qq.com>;"Mention"<mention@noreply.github.com>; 主题: Re: [ying09/TextFuseNet] how to make inferences (#18) No problem. Let me know if you face any other issue. I will be happy to help.😇 — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

Hi @StudentChen, yes you have to create json files in coco format.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

ying09 / TextFuseNet

how to make inferences #18