search

MichalBusta / E2E-MLT

E2E-MLT - an Unconstrained End-to-End Method for Multi-Language Scene Text
MIT License
291 stars 84 forks source link

about codec.txt #40

Open ustczhouyu opened 5 years ago

ustczhouyu commented 5 years ago

@MichalBusta Hi, there are 7398 characters in codec.txt, in the models.py, why self.conv11 = Conv2d(256, 8400, (1, 1), padding=(0,0))? I think it should be 7399

MichalBusta commented 5 years ago

Hi, there are 2 models: first for MLT 2017 challenge and second for MLT 2019 and RCTW challenge.

On Mon, 10 Jun 2019 11:56 ustczhouyu, notifications@github.com wrote:

@MichalBusta https://github.com/MichalBusta Hi, there are 7398 characters in codec.txt, in the models.py, why self.conv11 = Conv2d(256, 8400, (1, 1), padding=(0,0))? I think it should be 7399

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/MichalBusta/E2E-MLT/issues/40?email_source=notifications&email_token=AA7KHMFBXMVD7ARC2X7AIGTPZ2IVTA5CNFSM4HWVV3V2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GYT54VQ, or mute the thread https://github.com/notifications/unsubscribe-auth/AA7KHMGHNFPFCV64E2HX6A3PZ2IVTANCNFSM4HWVV3VQ .

ustczhouyu commented 5 years ago

Thanks for your reply. I know there are two models. 2017MLT codec.txt has 7398 characters, but conv11 has 7500 channels, maybe 7399 is more accurate, right?

zhou19920226

邮箱：zhou19920226@126.com |

Signature is customized by Netease Mail Master

On 06/13/2019 01:19, Michal Busta wrote: Hi, there are 2 models: first for MLT 2017 challenge and second for MLT 2019 and RCTW challenge.

On Mon, 10 Jun 2019 11:56 ustczhouyu, notifications@github.com wrote:

@MichalBusta https://github.com/MichalBusta Hi, there are 7398 characters in codec.txt, in the models.py, why self.conv11 = Conv2d(256, 8400, (1, 1), padding=(0,0))? I think it should be 7399

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/MichalBusta/E2E-MLT/issues/40?email_source=notifications&email_token=AA7KHMFBXMVD7ARC2X7AIGTPZ2IVTA5CNFSM4HWVV3V2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GYT54VQ, or mute the thread https://github.com/notifications/unsubscribe-auth/AA7KHMGHNFPFCV64E2HX6A3PZ2IVTANCNFSM4HWVV3VQ .

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

MichalBusta commented 5 years ago

Probably yes. There are also some very strange chars in codec (wide latin etc.) - it was generated without further inspection.

On Wed, 12 Jun 2019 11:24 ustczhouyu, notifications@github.com wrote:

Thanks for your reply. I know there are two models. 2017MLT codec.txt has 7398 characters, but conv11 has 7500 channels, maybe 7399 is more accurate, right?

zhou19920226

邮箱：zhou19920226@126.com |

Signature is customized by Netease Mail Master

On 06/13/2019 01:19, Michal Busta wrote: Hi, there are 2 models: first for MLT 2017 challenge and second for MLT 2019 and RCTW challenge.

On Mon, 10 Jun 2019 11:56 ustczhouyu, notifications@github.com wrote:

@MichalBusta https://github.com/MichalBusta Hi, there are 7398 characters in codec.txt, in the models.py, why self.conv11 = Conv2d(256, 8400, (1, 1), padding=(0,0))? I think it should be 7399

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub < https://github.com/MichalBusta/E2E-MLT/issues/40?email_source=notifications&email_token=AA7KHMFBXMVD7ARC2X7AIGTPZ2IVTA5CNFSM4HWVV3V2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GYT54VQ , or mute the thread < https://github.com/notifications/unsubscribe-auth/AA7KHMGHNFPFCV64E2HX6A3PZ2IVTANCNFSM4HWVV3VQ

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/MichalBusta/E2E-MLT/issues/40?email_source=notifications&email_token=AA7KHMD2ABK7X6SUQBWPGU3P2EWM5A5CNFSM4HWVV3V2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODXRF74Q#issuecomment-501374962, or mute the thread https://github.com/notifications/unsubscribe-auth/AA7KHMC7DGZOTUBQFO5IPIDP2EWM5ANCNFSM4HWVV3VQ .

ustczhouyu commented 5 years ago

Ok,thank you so much

zhou19920226

邮箱：zhou19920226@126.com |

Signature is customized by Netease Mail Master

On 06/13/2019 01:31, Michal Busta wrote: Probably yes. There are also some very strange chars in codec (wide latin etc.) - it was generated without further inspection.

On Wed, 12 Jun 2019 11:24 ustczhouyu, notifications@github.com wrote:

Thanks for your reply. I know there are two models. 2017MLT codec.txt has 7398 characters, but conv11 has 7500 channels, maybe 7399 is more accurate, right?

zhou19920226

邮箱：zhou19920226@126.com |

Signature is customized by Netease Mail Master

On 06/13/2019 01:19, Michal Busta wrote: Hi, there are 2 models: first for MLT 2017 challenge and second for MLT 2019 and RCTW challenge.

On Mon, 10 Jun 2019 11:56 ustczhouyu, notifications@github.com wrote:

@MichalBusta https://github.com/MichalBusta Hi, there are 7398 characters in codec.txt, in the models.py, why self.conv11 = Conv2d(256, 8400, (1, 1), padding=(0,0))? I think it should be 7399

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub < https://github.com/MichalBusta/E2E-MLT/issues/40?email_source=notifications&email_token=AA7KHMFBXMVD7ARC2X7AIGTPZ2IVTA5CNFSM4HWVV3V2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GYT54VQ , or mute the thread < https://github.com/notifications/unsubscribe-auth/AA7KHMGHNFPFCV64E2HX6A3PZ2IVTANCNFSM4HWVV3VQ

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/MichalBusta/E2E-MLT/issues/40?email_source=notifications&email_token=AA7KHMD2ABK7X6SUQBWPGU3P2EWM5A5CNFSM4HWVV3V2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODXRF74Q#issuecomment-501374962, or mute the thread https://github.com/notifications/unsubscribe-auth/AA7KHMC7DGZOTUBQFO5IPIDP2EWM5ANCNFSM4HWVV3VQ .

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.