You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+12-8Lines changed: 12 additions & 8 deletions
Original file line number
Diff line number
Diff line change
@@ -44,11 +44,11 @@ BEST PATH GPU : "the fak friend of the fomly hae tC"
44
44
* Token Passing: searches for most probable word sequence. The words are constrained to those contained in a dictionary. Can be extended to use a word-level LM. File: `TokenPassing.py`\[1\]
45
45
* Lexicon Search: computes approximation with best path decoding to find similar words in dictionary. Returns the one with highest score. File: `LexiconSearch.py`\[3\]
46
46
* Loss: calculates probability and loss of a given text in the RNN output. File: `Loss.py`\[1\]\[6\]
47
-
* Word Beam Search: TensorFlow implementation see repository [CTCWordBeamSearch](https://github.com/githubharald/CTCWordBeamSearch)
47
+
* Word Beam Search: TensorFlow implementation see repository [CTCWordBeamSearch](https://github.com/githubharald/CTCWordBeamSearch)\[8\]
48
48
49
49
50
50
## Choosing the right algorithm
51
-
[This paper](./doc/comparison.pdf) compares beam search decoding and token passing.
51
+
[This paper](./doc/comparison.pdf)\[7\]compares beam search decoding and token passing.
52
52
It gives suggestions when to use best path decoding, beam search decoding and token passing.
53
53
54
54
@@ -101,14 +101,18 @@ A GPU implementation is provided for best path decoding which requires pyopencl
101
101
102
102
## References
103
103
104
-
\[1\] Graves - Supervised sequence labelling with recurrent neural networks
104
+
\[1\][Graves - Supervised sequence labelling with recurrent neural networks](https://www.cs.toronto.edu/~graves/preprint.pdf)
\[2\][Hwang - Character-level incremental speech recognition with recurrent neural networks](https://arxiv.org/pdf/1601.06581.pdf)
107
107
108
-
\[3\] Shi - An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition: https://github.com/bgshih/crnn
108
+
\[3\][Shi - An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition](https://arxiv.org/pdf/1507.05717.pdf)
109
109
110
-
\[4\] Marti - The IAM-database: an English sentence database for offline handwriting recognition: http://www.fki.inf.unibe.ch/databases/iam-handwriting-database
110
+
\[4\][Marti - The IAM-database: an English sentence database for offline handwriting recognition](http://www.fki.inf.unibe.ch/databases/iam-handwriting-database)
111
111
112
-
\[5\] Beam Search Decoding in CTC-trained Neural Networks: https://towardsdatascience.com/5a889a3d85a7
112
+
\[5\][Beam Search Decoding in CTC-trained Neural Networks](https://towardsdatascience.com/5a889a3d85a7)
113
113
114
-
\[6\] An Intuitive Explanation of Connectionist Temporal Classification: https://towardsdatascience.com/3797e43a86c
114
+
\[6\][An Intuitive Explanation of Connectionist Temporal Classification](https://towardsdatascience.com/3797e43a86c)
115
+
116
+
\[7\][Scheidl - Comparison of Connectionist Temporal Classification Decoding Algorithms](./doc/comparison.pdf)
117
+
118
+
\[8\][Scheidl - Word Beam Search: A Connectionist Temporal Classification Decoding Algorithm](https://repositum.tuwien.ac.at/obvutwoa/download/pdf/2774578)
0 commit comments