Commits · 51fec0fdc6bcddfec4aa68a30f2f8676b573a7e4 · Fabian Mersch / SimpleHTR

Oct 23, 2023
- better preprocessor for training data · 51fec0fd
  merschie authored 1 year ago
  
  51fec0fd
Oct 11, 2023
- able to get all student infos · a5e6d89c
  fabian authored 1 year ago
  
  a5e6d89c
Oct 10, 2023
- Umlaute beachten · 8eeba10a
  fabian authored 1 year ago
  
  8eeba10a
Oct 06, 2023
- added lexicon search for Name recognition · 2e6282f0
  fabian authored 1 year ago
  
  2e6282f0
Aug 29, 2023
- updated for real exam example · abf00eb0
  fabian authored 1 year ago
  
  abf00eb0
Aug 18, 2023
- get top n beams · 1cd108db
  fabian authored 1 year ago
  
  1cd108db
Jul 12, 2023
- send all Matr. numbers · df06b7f5
  merschie authored 1 year ago
  
  df06b7f5
- split corpus into Vor und Nachname · 85721ca0
  merschie authored 1 year ago
  
  85721ca0
Jul 05, 2023
- add webserver · 1c614c73
  merschie authored 1 year ago
  
  1c614c73
- added model and corpus · 87fe2ec0
  merschie authored 1 year ago
  
  87fe2ec0
Apr 27, 2023
- added link to demo · 77d2a05b
  Harald Scheidl authored 2 years ago
  
  77d2a05b
Apr 03, 2023
- Merge pull request #157 from alexeevrevan/master · ec1ce163
  Harald Scheidl authored 2 years ago
  ec1ce163
Mar 29, 2023
- Fixed Bugs and Added Train Loss Dump Feature · 2b2a90a2
  Alexander Alexeev authored 2 years ago
  
  2b2a90a2
Mar 28, 2023
- added link to repo that uses this model · 2c14ec8e
  Harald Scheidl authored 2 years ago
  
  2c14ec8e
Oct 13, 2022
- ignore empty lines and comment lines in IAM dataset file · 8577a9e8
  Harald Scheidl authored 2 years ago
  
  8577a9e8
Mar 01, 2022
- in validation mode use charset from trained model and not from dataset, see issue #127 · d807bc44
  Harald Scheidl authored 3 years ago
  
  d807bc44
Jan 10, 2022
- png images: white instead of transparent bg to handle dark mode · 7f26b321
  Harald Scheidl authored 3 years ago
  
  7f26b321
May 27, 2021
- readme, docstrings, and other small changes · ab6220d6
  Harald Scheidl authored 4 years ago
  
  ab6220d6
May 25, 2021
- ready for first line-level training runs · 2b0b5484
  Harald Scheidl authored 4 years ago
  
  2b0b5484
May 24, 2021
- use correct batch size · dca7cabc
  Harald Scheidl authored 4 years ago
  
  dca7cabc
- reworked preprocessor · 3507e59d
  Harald Scheidl authored 4 years ago
  
  3507e59d
- started multi word training · 6c76bca2
  Harald Scheidl authored 4 years ago
  
  6c76bca2
- allow dynamic width of image · 7edd1f86
  Harald Scheidl authored 4 years ago
  
  7edd1f86
- changed code to python coding style · 8fd81536
  Harald Scheidl authored 4 years ago
  
  8fd81536
Feb 18, 2021
- rm old faq section · 8ba5e995
  Harald Scheidl authored 4 years ago
  
  8ba5e995
Feb 17, 2021
- word beam search: switched from TF custom op to Python package, so it can also... · a91b4632
  Harald Scheidl authored 4 years ago
  a91b4632
Feb 16, 2021
- cleanup: remove analyze code, and only keep code really relevant for HTR · 330c3dba
  Harald Scheidl authored 4 years ago
  
  330c3dba
Feb 03, 2021
- single option for decoders · a7e85ba7
  Harald Scheidl authored 4 years ago
  
  a7e85ba7
Feb 02, 2021
- log both word acc and cer for valset · 00d7232d
  Harald Scheidl authored 4 years ago
  
  00d7232d
- changed data aug · 8c96a80e
  Harald Scheidl authored 4 years ago
  
  8c96a80e
- data aug: dont move img around if it is already larger than target img · 5e191e35
  Harald Scheidl authored 4 years ago
  
  5e191e35
- all model files are now in the uploaded zip · 66a9f1b9
  Harald Scheidl authored 4 years ago
  
  66a9f1b9
- data aug: random background, write summary of CERs while training · cdfb5188
  Harald Scheidl authored 4 years ago
  
  cdfb5188
Feb 01, 2021
- test: data aug=true · 2f0499fc
  Harald Scheidl authored 4 years ago
  
  2f0499fc
- geometric data augmentation · ca133566
  Harald Scheidl authored 4 years ago
  
  ca133566
Jan 22, 2021
- update accuracy on full val set · 7bd8ca85
  Harald Scheidl authored 4 years ago
  
  7bd8ca85
- val set: always use all samples, even if last batch is smaller than specified batch size · 10781ce4
  Harald Scheidl authored 4 years ago
  
  10781ce4
- updated requirements · 339cceed
  Harald Scheidl authored 4 years ago
  
  339cceed
- reworked readme, optimized imports · 10366c39
  Harald Scheidl authored 4 years ago
  
  10366c39
Jan 21, 2021
- readme · 04e999f5
  Harald Scheidl authored 4 years ago
  
  04e999f5