Upgrade to KeenASR 1.62; new ASR Bundle; various cleanup

keenresearch · Jun 19, 2018 · 961e914 · 961e914
1 parent 9a84d7f
commit 961e914
Show file tree

Hide file tree

Showing 56 changed files with 769 additions and 896 deletions.
diff --git a/KeenASR.framework/VERSION.txt b/KeenASR.framework/VERSION.txt
@@ -1,5 +1,5 @@
-Version: 1.6
-Built on: Mon Jan 22 17:10:04 PST 2018
-Rev: 08f99777dc6faa9a948ffa42a3a2c79046e4f920
+Version: 1.62
+Built on: Sun Jun 17 21:28:25 PDT 2018
+Rev: 53508b3aa5d6ffd5cd106d8d87c93a9ccff843f4
 Bundle ID: THIS IS A TRIAL VERSION OF THE FRAMEWORK
 NOTE: these values are for reference only. They are not read by the framework.
diff --git a/KeenASR.framework/Versions/A/Headers/KIOSRecognizer.h b/KeenASR.framework/Versions/A/Headers/KIOSRecognizer.h
@@ -684,8 +684,8 @@ typedef NS_ENUM(NSInteger, KIOSVadParameter) {
  KeenASR audio stack if an audio interrupt comes through, and then
  reinit the audio stack when the interrupt is over. Setting handleNotifications
  to NO allows the SDK to work in the background mode; you will still need to
- properly handle audio interrupts using deactivateAudioSession,
- activateAudioSession, and stopListening methods.
+ properly handle audio interrupts using deactivateAudioStack,
+ activateAudioStack, and stopListening methods.
  */
 @property(nonatomic, assign, setter=setHandleNotifications:) BOOL handleNotifications;
 

diff --git a/KeenASR.framework/Versions/A/KeenASR b/KeenASR.framework/Versions/A/KeenASR
diff --git a/LICENSE b/LICENSE
diff --git a/README.md b/README.md
@@ -1,16 +1,20 @@
 ## Note
 
-This proof-of-concept app ships with a trial version of KeenASR framework, which will exit (crash) the app 10min after the framework has been initialized. If you would like to obtain a version of the framework without this limitation, contact us at info@keenresearch.com.
+This proof-of-concept app ships with a trial version of KeenASR framework, which will exit (crash) the app 15min after the framework has been initialized. If you would like to obtain a version of the framework without this limitation, contact us at info@keenresearch.com.
+
+By cloning this repository and downloading the trial KeenASR SDK or ASR Bundle you agree to the [KeenASR SDK Trial Licensing Agreement](https://keenresearch.com/keenasr-docs/keenasr-trial-sdk-licensing-agreement.html)
+
+For more details about the SDK see: http://keenresearch.com/keenasr-docs
 
 **Important:** 
 - You will need [git-lfs](https://git-lfs.github.com/) to checkout the project
-- After cloning the repository, you will need to **set/change the bundle id** for the app (currently set to com.keenresearch.com.kaldi-ios-poc), as well as **signing settings** in XCode project settings. These settings are under project build settings, General tab->Identity.
+- After cloning the repository, you will need to **set/change the bundle id** for the app (currently set to com.keenresearch.com.keenasr-ios-poc), as well as **signing settings** in XCode project settings. These settings are under project build settings, General tab->Identity.
 
 ## KeenASR Proof-of-Concept App
 
 A proof-of-concept app that shows how to run KeenASR automatic speech recognition framework. For detailed information on all classes and methods, consult the [SDK reference documentation](http://keenresearch.com/keenasr-docs). If starting with the framework from scratch, check our [Quick Start](http://keenresearch.com/keenasr-docs/docs/additional-docs/Quick-Start.html) document.
 
-This demo app uses acoustic models in librispeech-nnet2-en-us directory (originally from http://kaldi-asr.org/downloads/build/10/trunk/egs/librispeech/s5/exp/nnet2_online/nnet_ms_a_online). Keen Research provides a number of custom acoustic models to its clients.
+This demo app uses acoustic models in keenB2mQT-nnet3chain-en-us directory. Keen Research provides a number of custom acoustic models to its clients.
 
 Six different demos are provided in this proof of concept app:
 

diff --git a/keenB2mQT-nnet3chain-en-us/decode.conf b/keenB2mQT-nnet3chain-en-us/decode.conf
@@ -0,0 +1,13 @@
+--feature-type=mfcc
+--mfcc-config=keenB2mQT-nnet3chain-en-us/mfcc.conf
+--ivector-extraction-config=keenB2mQT-nnet3chain-en-us/ivector_extractor.conf
+--endpoint.silence-phones=1:2:3:4:5:6:7:8:9:10
+--frame-subsampling-factor=3
+--acoustic-scale=1
+--minimize=false
+--max-active=10000
+--min-active=500
+--beam=10
+--ivector-silence-weighting.silence-weight=0.001
+--ivector-silence-weighting.silence-phones=1:2:3:4:5:6:7:8:9:10
+--ivector-silence-weighting.max-state-duration=40
diff --git a/keenB2mQT-nnet3chain-en-us/final.dubm b/keenB2mQT-nnet3chain-en-us/final.dubm
diff --git a/librispeechQT-nnet2-en-us/final.ie → keenB2mQT-nnet3chain-en-us/final.ie b/librispeechQT-nnet2-en-us/final.ie → keenB2mQT-nnet3chain-en-us/final.ie
diff --git a/keenB2mQT-nnet3chain-en-us/final.mat b/keenB2mQT-nnet3chain-en-us/final.mat
diff --git a/keenB2mQT-nnet3chain-en-us/final.mdl b/keenB2mQT-nnet3chain-en-us/final.mdl
diff --git a/keenB2mQT-nnet3chain-en-us/global_cmvn.stats b/keenB2mQT-nnet3chain-en-us/global_cmvn.stats
@@ -0,0 +1,3 @@
+ [
+  1.112302e+11 -1.041458e+10 -1.1052e+10 -3.707e+08 -1.585259e+10 -1.264665e+10 -1.583232e+10 -1.00904e+10 -6.383481e+09 -5.328522e+09 -5.280869e+09 -5.870139e+09 -4.895705e+09 -4.880152e+09 -6.342072e+09 -3.34958e+09 -3.259781e+09 -1.336062e+09 -9.541146e+08 -4.130515e+08 -1.109862e+08 -1.181156e+08 -5927524 -1.930181e+07 8.851942e+07 8.50938e+07 8.541277e+07 8.869975e+07 -5.631357e+07 3.437372e+07 -3.646938e+07 1.910262e+08 3.645957e+08 5.281024e+08 5.756281e+08 3.979867e+08 2.525087e+08 -3.231467e+08 -3.477414e+08 -2.382089e+08 1.161776e+09 
+  1.10864e+13 6.831981e+11 5.76054e+11 5.535469e+11 8.01044e+11 7.486306e+11 7.884192e+11 6.505788e+11 5.507396e+11 5.045179e+11 4.633392e+11 4.404773e+11 3.536921e+11 3.378102e+11 2.868802e+11 2.06073e+11 1.630016e+11 1.151408e+11 7.523919e+10 4.412853e+10 2.35141e+10 8.013048e+09 1.033368e+09 2.975412e+08 3.578637e+09 9.178016e+09 1.551411e+10 2.264768e+10 2.694816e+10 2.980657e+10 3.089701e+10 3.215522e+10 3.0984e+10 3.248089e+10 2.61353e+10 2.093135e+10 1.985831e+10 1.649075e+10 1.219972e+10 8.42584e+09 0 ]
diff --git a/keenB2mQT-nnet3chain-en-us/ivector_extractor.conf b/keenB2mQT-nnet3chain-en-us/ivector_extractor.conf
@@ -0,0 +1,11 @@
+--splice-config=keenB2mQT-nnet3chain-en-us/splice.conf
+--cmvn-config=keenB2mQT-nnet3chain-en-us/online_cmvn.conf
+--lda-matrix=keenB2mQT-nnet3chain-en-us/final.mat
+--global-cmvn-stats=keenB2mQT-nnet3chain-en-us/global_cmvn.stats
+--diag-ubm=keenB2mQT-nnet3chain-en-us/final.dubm
+--ivector-extractor=keenB2mQT-nnet3chain-en-us/final.ie
+--num-gselect=5
+--min-post=0.025
+--posterior-scale=0.1
+--max-remembered-frames=1000
+--max-count=100
diff --git a/librispeechQT-nnet2-en-us/lang/lexicon.txt → keenB2mQT-nnet3chain-en-us/lang/lexicon.txt b/librispeechQT-nnet2-en-us/lang/lexicon.txt → keenB2mQT-nnet3chain-en-us/lang/lexicon.txt
@@ -1,6 +1,5 @@
-SIL	SIL_S
-<SPOKEN_NOISE>	SPN_S
-<UNK>	SPN_S
+!SIL  SIL_S
+<SPOKEN_NOISE>  SPN_S
 A	EY1_S
 A	AH0_S
 A''S	EY1_B	Z_E
@@ -19031,6 +19030,7 @@ BLOODIER	B_B	L_I	AH1_I	D_I	IY0_I	ER0_E
 BLOODIEST	B_B	L_I	AH1_I	D_I	IY0_I	AH0_I	S_I	T_E
 BLOODILY	B_B	L_I	AH1_I	D_I	AH0_I	L_I	IY0_E
 BLOODING	B_B	L_I	AH1_I	D_I	IH0_I	NG_E
+BLOODLINE	B_B	L_I	AH1_I	D_I	L_I	AY0_I	N_E
 BLOODLESS	B_B	L_I	AH1_I	D_I	L_I	AH0_I	S_E
 BLOODLESSLY	B_B	L_I	AH1_I	D_I	L_I	AH0_I	S_I	L_I	IY0_E
 BLOODLESSNESS	B_B	L_I	AH1_I	D_I	L_I	AH0_I	S_I	N_I	AH0_I	S_E