Please enable JavaScript.

Coggle requires JavaScript to display documents.

Contribution of recurrent connectionist language models in improving …

- - - - In 1914, Emanuel Goldberg
      - Techniques
        
        Pre-processing
        
        Character recognition
        
        Post-processing
  - - - bidirectional LSTM-CTC
        
        CTC（Connectionist Temporal Classification，連續時間序列分類）
    - - 最大熵模糊
    - - 集束搜索演算法http://www.pianshen.com/article/369392933/
    - - 一般來說OCR 字符識別技術可分為
        
        Feature extraction
        
        Matrix matching
  - - - The field of pattern recognition is concerned with the automatic discovery of regularities in data through the use of computer algorithms and with the use of these regularities to take actions such as classifying the data into different categories.
  - - - sequence labeling, which assigns a class to each member of a sequence of values
      - Sequence labeling algorithms (predicting sequences of categorical labels)
        
        Unsupervised:
        
        Hidden Markov models (HMMs)
        
        Dynamic time warping (DTW)
        
        Supervised:
        
        Conditional random fields (CRFs)
        
        Hidden Markov models (HMMs)
        
        Maximum entropy Markov models (MEMMs)
        
        Recurrent neural networks (RNNs)
  - - - CNN+RNN+CTC
        
        https://arxiv.org/abs/1507.05717
      - CNN+RNN 基於 Attention
        
        http://papers.nips.cc/paper/5346-sequence-to-sequence-learning-with-neural-networks.pdf
- - - - CTC 是一種技術,能針對每一個輸入的序列輸出一個標籤或是空白
    - - 統計方法由 n-1的字詞判斷下一個詞出現的的機率