datapath = '' modelpath = 'model_speech' system_type = plat.system() # 由于不同的系统的文件路径表示不一样,需要进行判断 if (system_type == 'Windows'): datapath = 'E:\\语音数据集' modelpath = modelpath + '\\' elif (system_type == 'Linux'): datapath = 'dataset' modelpath = modelpath + '/' else: print('*[Message] Unknown System\n') datapath = 'dataset' modelpath = modelpath + '/' ms = ModelSpeech(datapath) ms.LoadModel(modelpath + 'm_dfcnn/speech_model_dfcnn_e_0_step_' + EPOCH + '000.model') #ms.LoadModel(modelpath + 'm_DFCNN/speech_model_DFCNN_e_0_step_8000.model') #ms.LoadModel(modelpath + 'm26/speech_model26_e_0_step_122500.model') #ms.TestModel(datapath, str_dataset='test', data_count = 64, out_report = True) #r = ms.RecognizeSpeech_FromFile('myspeech.wav') #r = ms.RecognizeSpeech_FromFile('/home/speech.AI/github/DFCNN/dataset/data_thchs30/test/D11_750.wav') #aishell r = ms.RecognizeSpeech_FromFile( '/home/speech.AI/github/DFCNN/dataset/data_aishell/wav/test/S0764/BAC009S0764W0122.wav' ) #r = ms.RecognizeSpeech_FromFile('/home/speech.AI/github/DFCNN/dataset/data_aishell/wav/test/S0764/BAC009S0764W0124.wav') #r = ms.RecognizeSpeech_FromFile('/home/speech.AI/github/DFCNN/dataset/data_aishell/wav/test/S0764/BAC009S0764W0179.wav')
#!/usr/bin/env python3 # -*- coding: utf-8 -*- """ 语音识别API的HTTP服务器程序 """ import http.server import urllib import keras from SpeechModel_DFCNN import ModelSpeech from LanguageModel import ModelLanguage datapath = 'data/' modelpath = 'model_speech/' ms = ModelSpeech(datapath) ms.LoadModel(modelpath + 'm_DFCNN/speech_model_DFCNN_e_0_step_410000.model') ml = ModelLanguage('model_language') ml.LoadModel() class TestHTTPHandle(http.server.BaseHTTPRequestHandler): def setup(self): self.request.settimeout(10) http.server.BaseHTTPRequestHandler.setup(self) def _set_response(self): self.send_response(200) self.send_header('Content-type', 'text/html') self.end_headers()
datapath = '' modelpath = 'model_speech' f = open('step_dfcnn.txt', 'r') flist = f.readlines() f.close() fstr = "".join(flist) base_count = fstr.split('_')[-1] if (not os.path.exists(modelpath)): # 判断保存模型的目录是否存在 os.makedirs(modelpath) # 如果不存在,就新建一个,避免之后保存模型的时候炸掉 system_type = plat.system() # 由于不同的系统的文件路径表示不一样,需要进行判断 if (system_type == 'Windows'): datapath = 'E:\\语音数据集' modelpath = modelpath + '\\' elif (system_type == 'Linux'): datapath = 'dataset' modelpath = modelpath + '/' else: print('*[Message] Unknown System\n') datapath = 'dataset' modelpath = modelpath + '/' ms = ModelSpeech(datapath) #ms.LoadModel(modelpath + 'm_DFCNN/speech_model_DFCNN_e_0_step_84000.model') #ms.LoadModel(modelpath + 'speech_model_dfcnn_e_0_step_' + base_count +'.model') ms.TrainModel(datapath, epoch=50, batch_size=16, save_step=1000)
modelpath = 'model_speech' f = open('step_dfcnn.txt','r') flist = f.readlines() f.close() fstr = "".join(flist) base_count=fstr.split('_')[-1] if(not os.path.exists(modelpath)): # 判断保存模型的目录是否存在 os.makedirs(modelpath) # 如果不存在,就新建一个,避免之后保存模型的时候炸掉 system_type = plat.system() # 由于不同的系统的文件路径表示不一样,需要进行判断 if(system_type == 'Windows'): datapath = 'E:\\语音数据集' modelpath = modelpath + '\\' elif(system_type == 'Linux'): datapath = 'dataset' modelpath = modelpath + '/' else: print('*[Message] Unknown System\n') datapath = 'dataset' modelpath = modelpath + '/' ms = ModelSpeech(datapath) #ms.LoadModel(modelpath + 'm_DFCNN/speech_model_DFCNN_e_0_step_84000.model') ms.LoadModel(modelpath + 'm_dfcnn/speech_model_dfcnn_e_0_step_' + base_count +'.model') ms.TrainModel(datapath, epoch = 50, batch_size = 64, save_step = 1000)
config = tf.ConfigProto() config.gpu_options.per_process_gpu_memory_fraction = 0.9 #config.gpu_options.allow_growth=True #不全部占满显存, 按需分配 set_session(tf.Session(config=config)) ''' os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID" os.environ["CUDA_VISIBLE_DEVICES"] = "0, 1" #进行配置,使用70%的GPU config = tf.ConfigProto(allow_soft_placement=True) config.gpu_options.per_process_gpu_memory_fraction = 0.95 #config.gpu_options.allow_growth=True #不全部占满显存, 按需分配 set_session(tf.Session(config=config)) datapath = '' modelpath = 'model_speech/' if (not os.path.exists(modelpath)): # 判断保存模型的目录是否存在 os.makedirs(modelpath) # 如果不存在,就新建一个,避免之后保存模型的时候炸掉 ms = ModelSpeech(datapath) ms.LoadModel(modelpath + 'm_dfcnn/speech_model_dfcnn_e_0_step_64000.model') #ms.LoadModel(modelpath + 'm_DFCNN/speech_model_DFCNN_e_0_step_410000.model') ms.TestModel(datapath, str_dataset='test', data_count=64, out_report=True) #r = ms.RecognizeSpeech_FromFile('E:\\语音数据集\\ST-CMDS-20170001_1-OS\\20170001P00241I0053.wav') #r = ms.RecognizeSpeech_FromFile('E:\\语音数据集\\ST-CMDS-20170001_1-OS\\20170001P00020I0087.wav') #r = ms.RecognizeSpeech_FromFile('E:\\语音数据集\\wav\\train\\A11\\A11_167.WAV') #r = ms.RecognizeSpeech_FromFile('E:\\语音数据集\\wav\\test\\D4\\D4_750.wav') #print('*[提示] 语音识别结果:\n',r)
from keras.backend.tensorflow_backend import set_session from SpeechModel_DFCNN import ModelSpeech from LanguageModel import ModelLanguage os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID" os.environ["CUDA_VISIBLE_DEVICES"] = "0, 1" #进行配置,使用70%的GPU config = tf.ConfigProto(allow_soft_placement=True) config.gpu_options.per_process_gpu_memory_fraction = 0.9 #config.gpu_options.allow_growth=True #不全部占满显存, 按需分配 set_session(tf.Session(config=config)) datapath = '' modelpath = 'model_speech/' ms = ModelSpeech(datapath) ms.LoadModel(modelpath + 'm_dfcnn/speech_model_dfcnn_e_0_step_64000.model') #ms.LoadModel(modelpath + 'm_DFCNN/speech_model_DFCNN_e_0_step_410000.model') #ms.LoadModel(modelpath + 'm26/speech_model26_e_0_step_122500.model') #ms.TestModel(datapath, str_dataset='test', data_count = 64, out_report = True) #r = ms.RecognizeSpeech_FromFile('/home/speech.AI/github/DFCNN/dataset/data_thchs30/test/D11_750.wav') r = ms.RecognizeSpeech_FromFile( '/home/speech.AI/github/DFCNN/dataset/data_thchs30/train/A33_100.wav') print('*[提示] 语音识别结果:\n', r) ml = ModelLanguage('model_language') ml.LoadModel() #str_pinyin = ['zhe4','zhen1','shi4','ji2', 'hao3','de5']
if(not os.path.exists(modelpath)): # 判断保存模型的目录是否存在 os.makedirs(modelpath) # 如果不存在,就新建一个,避免之后保存模型的时候炸掉 system_type = plat.system() # 由于不同的系统的文件路径表示不一样,需要进行判断 if(system_type == 'Windows'): datapath = 'E:\\语音数据集' modelpath = modelpath + '\\' elif(system_type == 'Linux'): datapath = 'dataset' modelpath = modelpath + '/' else: print('*[Message] Unknown System\n') datapath = 'dataset' modelpath = modelpath + '/' ms = ModelSpeech(datapath) ms.LoadModel(modelpath + 'm_dfcnn/speech_model_dfcnn_e_0_step_'+iters_num+'000.model') #ms.LoadModel(modelpath + 'm_DFCNN/speech_model_DFCNN_e_0_step_84000.model') ms.TestModel(datapath, str_dataset='test', data_count = 64, out_report = True) r = ms.RecognizeSpeech_FromFile('dataset/data_aishell/wav/dev/S0733/BAC009S0733W0234.wav') #r = ms.RecognizeSpeech_FromFile('E:\\语音数据集\\ST-CMDS-20170001_1-OS\\20170001P00020I0087.wav') #r = ms.RecognizeSpeech_FromFile('E:\\语音数据集\\wav\\train\\A11\\A11_167.WAV') #r = ms.RecognizeSpeech_FromFile('E:\\语音数据集\\wav\\test\\D4\\D4_750.wav') print('*[提示] 语音识别结果:\n',r)