w0rdplay

所属分类:远程控制/远程桌面
开发工具:Python
文件大小:43248KB
下载次数:0
上传日期:2018-02-02 20:07:58
上 传 者sh-1993
说明:  发音分类器和字符串拆分器
(Pronounceability classifier and string splitter)

文件列表:
analyze.py (1104, 2016-10-09)
cmuarctic_g2p.lua (2454, 2016-10-09)
data (0, 2016-10-09)
data\cmuarctic (0, 2016-10-09)
data\cmuarctic\cmuarctic.data.txt (77059, 2016-10-09)
data\cmuarctic\data_with_phonemes.txt (207262, 2016-10-09)
data\garbage (0, 2016-10-09)
data\garbage\garbage.txt (31, 2016-10-09)
data\garbage\good.txt (78, 2016-10-09)
data\garbage\letter_bigram_counts.txt (9269, 2016-10-09)
data\logs (0, 2016-10-09)
data\logs\logs.csv (722274, 2016-10-09)
data\logs\spoken_words.csv (46854, 2016-10-09)
data\logs\spoken_words.pkl (367758, 2016-10-09)
data\logs\spoken_words_all.csv (69041, 2016-10-09)
data\word_freqs (0, 2016-10-09)
data\word_freqs\google_10k_words_by_freq.txt (75879, 2016-10-09)
data\word_freqs\wict_100k_words_by_freq.txt (823537, 2016-10-09)
dl_wiki_subset.sh (124, 2016-10-09)
g2p (0, 2016-10-09)
g2p\MaskRNN.lua (2011, 2016-10-09)
g2p\ReverseMaskRNN.lua (2483, 2016-10-09)
g2p\g2p.t7 (63964316, 2016-10-09)
g2p\g2p_utils.lua (2113, 2016-10-09)
g2p\grapheme_to_idx.t7 (558, 2016-10-09)
g2p\lua_utils.lua (6416, 2016-10-09)
g2p\phoneme_to_idx.t7 (854, 2016-10-09)
garbage_detector.py (4864, 2016-10-09)
gui (0, 2016-10-09)
gui\.DS_Store (6148, 2016-10-09)
gui\public (0, 2016-10-09)
gui\public\.DS_Store (6148, 2016-10-09)
gui\public\css (0, 2016-10-09)
gui\public\css\base.css (1346, 2016-10-09)
gui\public\index.html (1013, 2016-10-09)
gui\public\scripts (0, 2016-10-09)
gui\public\scripts\example.js (7199, 2016-10-09)
... ...

# w0rdplay Modules to: 1. help analyze Speechblocks log data (e.g. detecting intentionality of word construction) 2. help build the text analysis portion of a speech synthesizer ## Garbage Word Detector ## Detect words that are not pronounceable, e.g. xlkokm ## String Splitter ## Split a string into words, e.g. lionxballcatyu -> lion x ball cat yu ## Running ## A front-end is availble to explore some of the modules. Currently, you can type words and have them classified as pronounceable or not (green is pronounceable, red is not). The classification happens dynamically as you type. String splitting into words occurs once you hit enter. You can also click on a list of words from Speechblocks log data, which are all the words that were pronounced during the pilot study. Get python packages using requirements.txt Download word frequency list by running: ``` ./dl_wiki_subset.sh ``` Start server by running: ``` PYTHONPATH=. python gui/server.py ``` Visit: ``` localhost:3000 ```

近期下载者

相关文件


收藏者