22d1e4ce10
2024.01.01: openai token计算✓ 优化ReAct Agent 与 LLM Chain自动切换逻辑✓ *添加双记忆机制：长时记忆流及短时聊天记忆✓ 修复record.py asr bug✓ 提高远程音频（android 连接器）的稳定性✓ 修复执行时间计算bug✓ 优化语音输出逻辑✓
349 lines
14 KiB
Python
349 lines
14 KiB
Python
|
||
import math
|
||
import os
|
||
import time
|
||
import socket
|
||
|
||
import eyed3
|
||
import logging
|
||
|
||
|
||
# 适应模型使用
|
||
import numpy as np
|
||
import fay_booter
|
||
from ai_module import xf_ltp
|
||
from ai_module.openai_tts import Speech
|
||
from core import wsa_server
|
||
from core.interact import Interact
|
||
from scheduler.thread_manager import MyThread
|
||
from utils import util, config_util
|
||
|
||
import pygame
|
||
from utils import config_util as cfg
|
||
from ai_module import nlp_cemotion
|
||
import platform
|
||
from ai_module import yolov8
|
||
from agent import agent_service
|
||
import fay_booter
|
||
from core.content_db import Content_Db
|
||
if platform.system() == "Windows":
|
||
import sys
|
||
sys.path.append("test/ovr_lipsync")
|
||
from test_olipsync import LipSyncGenerator
|
||
|
||
|
||
#文本消息处理（20231211增加：agent操作）
|
||
def send_for_answer(msg):
|
||
#记录运行时间
|
||
fay_booter.feiFei.last_quest_time = time.time()
|
||
|
||
#消息保存
|
||
contentdb = Content_Db()
|
||
contentdb.add_content('member', 'agent', msg.replace('主人语音说了：', '').replace('主人文字说了：', ''))
|
||
wsa_server.get_web_instance().add_cmd({"panelReply": {"type":"member","content":msg.replace('主人语音说了：', '').replace('主人文字说了：', '')}})
|
||
|
||
# 发送给数字人端
|
||
if not config_util.config["interact"]["playSound"]:
|
||
content = {'Topic': 'Unreal', 'Data': {'Key': 'question', 'Value': msg}}
|
||
wsa_server.get_instance().add_cmd(content)
|
||
|
||
#思考中...
|
||
wsa_server.get_web_instance().add_cmd({"panelMsg": "思考中..."})
|
||
if not cfg.config["interact"]["playSound"]: # 非展板播放
|
||
content = {'Topic': 'Unreal', 'Data': {'Key': 'log', 'Value': "思考中..."}}
|
||
wsa_server.get_instance().add_cmd(content)
|
||
|
||
#agent 或llm chain处理
|
||
is_use_say_tool, text = agent_service.agent.run(msg)
|
||
|
||
#语音输入强制语音输出
|
||
if text and "语音说了" in msg and not is_use_say_tool:
|
||
interact = Interact("audio", 1, {'user': '', 'msg': text})
|
||
fay_booter.feiFei.on_interact(interact)
|
||
|
||
#消息保存
|
||
contentdb.add_content('fay','agent', text)
|
||
wsa_server.get_web_instance().add_cmd({"panelReply": {"type":"fay","content":text}})
|
||
util.log(1, 'ReAct Agent或LLM Chain处理总时长：{} ms'.format(math.floor((time.time() - fay_booter.feiFei.last_quest_time) * 1000)))
|
||
|
||
#推送数字人
|
||
if not cfg.config["interact"]["playSound"]:
|
||
content = {'Topic': 'Unreal', 'Data': {'Key': 'log', 'Value': text}}
|
||
wsa_server.get_instance().add_cmd(content)
|
||
if not config_util.config["interact"]["playSound"]:
|
||
content = {'Topic': 'Unreal', 'Data': {'Key': 'text', 'Value': text}}
|
||
wsa_server.get_instance().add_cmd(content)
|
||
|
||
return text
|
||
|
||
|
||
class FeiFei:
|
||
def __init__(self):
|
||
pygame.mixer.init()
|
||
self.q_msg = '你叫什么名字？'
|
||
self.a_msg = 'hi,我叫菲菲，英文名是fay'
|
||
self.mood = 0.0 # 情绪值
|
||
self.old_mood = 0.0
|
||
self.connect = False
|
||
self.item_index = 0
|
||
self.deviceSocket = None
|
||
self.deviceConnect = None
|
||
|
||
#启动音频输入输出设备的连接服务
|
||
self.deviceSocketThread = MyThread(target=self.__accept_audio_device_output_connect)
|
||
self.deviceSocketThread.start()
|
||
|
||
self.X = np.array([1, 0, 0, 0, 0, 0, 0, 0]).reshape(1, -1) # 适应模型变量矩阵
|
||
# self.W = np.array([0.01577594,1.16119452,0.75828,0.207746,1.25017864,0.1044121,0.4294899,0.2770932]).reshape(-1,1) #适应模型变量矩阵
|
||
self.W = np.array([0.0, 0.6, 0.1, 0.7, 0.3, 0.0, 0.0, 0.0]).reshape(-1, 1) # 适应模型变量矩阵
|
||
|
||
self.wsParam = None
|
||
self.wss = None
|
||
self.sp = Speech()
|
||
self.speaking = False
|
||
self.interactive = []
|
||
self.sleep = False
|
||
self.__running = True
|
||
self.sp.connect() # 预连接
|
||
self.last_quest_time = time.time()
|
||
self.playing = False
|
||
self.muting = False
|
||
self.cemotion = None
|
||
|
||
|
||
def __auto_speak(self):
|
||
while self.__running:
|
||
time.sleep(0.1)
|
||
if self.speaking or self.sleep:
|
||
continue
|
||
|
||
try:
|
||
if len(self.interactive) > 0:
|
||
interact: Interact = self.interactive.pop()
|
||
#开启fay eyes，无人时不回复
|
||
fay_eyes = yolov8.new_instance()
|
||
if fay_eyes.get_status():#YOLO正在运行
|
||
person_count, stand_count, sit_count = fay_eyes.get_counts()
|
||
if person_count < 1: #看不到人，不互动
|
||
wsa_server.get_web_instance().add_cmd({"panelMsg": "看不到人，不互动"})
|
||
if not cfg.config["interact"]["playSound"]: # 非展板播放
|
||
content = {'Topic': 'Unreal', 'Data': {'Key': 'log', 'Value': "看不到人，不互动"}}
|
||
wsa_server.get_instance().add_cmd(content)
|
||
continue
|
||
|
||
self.speaking = True
|
||
self.a_msg = interact.data["msg"]
|
||
MyThread(target=self.__say, args=['interact']).start()
|
||
|
||
except BaseException as e:
|
||
print(e)
|
||
|
||
def on_interact(self, interact: Interact):
|
||
self.interactive.append(interact)
|
||
MyThread(target=self.__update_mood, args=[interact.interact_type]).start()
|
||
|
||
|
||
# 适应模型计算(用于学习真人的性格特质，开源版本暂不使用)
|
||
def __fay(self, index):
|
||
if 0 < index < 8:
|
||
self.X[0][index] += 1
|
||
# PRED = 1 /(1 + tf.exp(-tf.matmul(tf.constant(self.X,tf.float32), tf.constant(self.W,tf.float32))))
|
||
PRED = np.sum(self.X.reshape(-1) * self.W.reshape(-1))
|
||
if 0 < index < 8:
|
||
print('***PRED:{0}***'.format(PRED))
|
||
print(self.X.reshape(-1) * self.W.reshape(-1))
|
||
return PRED
|
||
|
||
# 发送情绪
|
||
def __send_mood(self):
|
||
while self.__running:
|
||
time.sleep(3)
|
||
if not self.sleep and not config_util.config["interact"]["playSound"] and wsa_server.get_instance().isConnect:
|
||
content = {'Topic': 'Unreal', 'Data': {'Key': 'mood', 'Value': self.mood}}
|
||
if not self.connect:
|
||
wsa_server.get_instance().add_cmd(content)
|
||
self.connect = True
|
||
else:
|
||
if self.old_mood != self.mood:
|
||
wsa_server.get_instance().add_cmd(content)
|
||
self.old_mood = self.mood
|
||
|
||
else:
|
||
self.connect = False
|
||
|
||
# 更新情绪
|
||
def __update_mood(self, typeIndex):
|
||
perception = config_util.config["interact"]["perception"]
|
||
if typeIndex == 1:
|
||
try:
|
||
if cfg.ltp_mode == "cemotion":
|
||
result = nlp_cemotion.get_sentiment(self.cemotion,self.q_msg)
|
||
chat_perception = perception["chat"]
|
||
if result >= 0.5 and result <= 1:
|
||
self.mood = self.mood + (chat_perception / 200.0)
|
||
elif result <= 0.2:
|
||
self.mood = self.mood - (chat_perception / 100.0)
|
||
else:
|
||
result = xf_ltp.get_sentiment(self.q_msg)
|
||
chat_perception = perception["chat"]
|
||
if result == 1:
|
||
self.mood = self.mood + (chat_perception / 200.0)
|
||
elif result == -1:
|
||
self.mood = self.mood - (chat_perception / 100.0)
|
||
except BaseException as e:
|
||
print("[System] 情绪更新错误！")
|
||
print(e)
|
||
|
||
elif typeIndex == 2:
|
||
self.mood = self.mood + (perception["join"] / 100.0)
|
||
|
||
elif typeIndex == 3:
|
||
self.mood = self.mood + (perception["gift"] / 100.0)
|
||
|
||
elif typeIndex == 4:
|
||
self.mood = self.mood + (perception["follow"] / 100.0)
|
||
|
||
if self.mood >= 1:
|
||
self.mood = 1
|
||
if self.mood <= -1:
|
||
self.mood = -1
|
||
|
||
def __get_mood_voice(self):
|
||
voice = config_util.config["attribute"]["voice"]
|
||
return voice
|
||
|
||
# 合成声音
|
||
def __say(self, styleType):
|
||
try:
|
||
if len(self.a_msg) < 1:
|
||
self.speaking = False
|
||
else:
|
||
util.printInfo(1, '菲菲', '({}) {}'.format(self.__get_mood_voice(), self.a_msg))
|
||
util.log(1, '合成音频...')
|
||
tm = time.time()
|
||
#文字也推送出去，为了ue5
|
||
result = self.sp.to_sample(self.a_msg, self.__get_mood_voice())
|
||
util.log(1, '合成音频完成. 耗时: {} ms 文件:{}'.format(math.floor((time.time() - tm) * 1000), result))
|
||
if result is not None:
|
||
MyThread(target=self.__send_or_play_audio, args=[result, styleType]).start()
|
||
return result
|
||
except BaseException as e:
|
||
print(e)
|
||
self.speaking = False
|
||
return None
|
||
|
||
def __play_sound(self, file_url):
|
||
util.log(1, '播放音频...')
|
||
util.log(1, 'agent处理总时长：{} ms'.format(math.floor((time.time() - self.last_quest_time) * 1000)))
|
||
pygame.mixer.music.load(file_url)
|
||
pygame.mixer.music.play()
|
||
|
||
|
||
def __send_or_play_audio(self, file_url, say_type):
|
||
try:
|
||
try:
|
||
logging.getLogger('eyed3').setLevel(logging.ERROR)
|
||
audio_length = eyed3.load(file_url).info.time_secs #mp3音频长度
|
||
except Exception as e:
|
||
audio_length = 3
|
||
|
||
# with wave.open(file_url, 'rb') as wav_file: #wav音频长度
|
||
# audio_length = wav_file.getnframes() / float(wav_file.getframerate())
|
||
# print(audio_length)
|
||
# if audio_length <= config_util.config["interact"]["maxInteractTime"] or say_type == "script":
|
||
if config_util.config["interact"]["playSound"]: # 展板播放
|
||
self.__play_sound(file_url)
|
||
else:#发送音频给ue
|
||
#推送ue
|
||
content = {'Topic': 'Unreal', 'Data': {'Key': 'audio', 'Value': os.path.abspath(file_url), 'Text': self.a_msg, 'Time': audio_length, 'Type': say_type}}
|
||
#计算lips
|
||
if platform.system() == "Windows":
|
||
try:
|
||
lip_sync_generator = LipSyncGenerator()
|
||
viseme_list = lip_sync_generator.generate_visemes(os.path.abspath(file_url))
|
||
consolidated_visemes = lip_sync_generator.consolidate_visemes(viseme_list)
|
||
content["Data"]["Lips"] = consolidated_visemes
|
||
except Exception as e:
|
||
util.log(1, "唇型数字生成失败，无法使用新版ue5工程")
|
||
wsa_server.get_instance().add_cmd(content)
|
||
|
||
#推送远程音频
|
||
if self.deviceConnect is not None:
|
||
try:
|
||
self.deviceConnect.send(b'\x00\x01\x02\x03\x04\x05\x06\x07\x08') # 发送音频开始标志，同时也检查设备是否在线
|
||
wavfile = open(os.path.abspath(file_url),'rb')
|
||
data = wavfile.read(1024)
|
||
total = 0
|
||
while data:
|
||
total += len(data)
|
||
self.deviceConnect.send(data)
|
||
data = wavfile.read(1024)
|
||
time.sleep(0.001)
|
||
self.deviceConnect.send(b'\x08\x07\x06\x05\x04\x03\x02\x01\x00')# 发送音频结束标志
|
||
util.log(1, "远程音频发送完成：{}".format(total))
|
||
except socket.error as serr:
|
||
util.log(1,"远程音频输入输出设备已经断开：{}".format(serr))
|
||
|
||
time.sleep(audio_length + 0.5)
|
||
wsa_server.get_web_instance().add_cmd({"panelMsg": ""})
|
||
if not cfg.config["interact"]["playSound"]: # 非展板播放
|
||
content = {'Topic': 'Unreal', 'Data': {'Key': 'log', 'Value': ""}}
|
||
wsa_server.get_instance().add_cmd(content)
|
||
if config_util.config["interact"]["playSound"]:
|
||
util.log(1, '结束播放！')
|
||
self.speaking = False
|
||
except Exception as e:
|
||
print(e)
|
||
|
||
def __device_socket_keep_alive(self):
|
||
while True:
|
||
if self.deviceConnect is not None:
|
||
try:
|
||
self.deviceConnect.send(b'\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xf7\xf8')#发送心跳包
|
||
except Exception as serr:
|
||
util.log(1,"远程音频输入输出设备已经断开：{}".format(serr))
|
||
self.deviceConnect = None
|
||
time.sleep(5)
|
||
|
||
def __accept_audio_device_output_connect(self):
|
||
self.deviceSocket = socket.socket(socket.AF_INET,socket.SOCK_STREAM)
|
||
self.deviceSocket.bind(("0.0.0.0",10001))
|
||
self.deviceSocket.listen(1)
|
||
addr = None
|
||
try:
|
||
while True:
|
||
self.deviceConnect,addr=self.deviceSocket.accept() #接受TCP连接，并返回新的套接字与IP地址
|
||
MyThread(target=self.__device_socket_keep_alive).start() # 开启心跳包检测
|
||
util.log(1,"远程音频输入输出设备连接上：{}".format(addr))
|
||
while self.deviceConnect: #只允许一个设备连接
|
||
time.sleep(1)
|
||
except Exception as err:
|
||
pass
|
||
|
||
def set_sleep(self, sleep):
|
||
self.sleep = sleep
|
||
|
||
def start(self):
|
||
if cfg.ltp_mode == "cemotion":
|
||
from cemotion import Cemotion
|
||
self.cemotion = Cemotion()
|
||
MyThread(target=self.__send_mood).start()
|
||
MyThread(target=self.__auto_speak).start()
|
||
|
||
|
||
def stop(self):
|
||
self.__running = False
|
||
self.speaking = False
|
||
self.playing = False
|
||
self.sp.close()
|
||
wsa_server.get_web_instance().add_cmd({"panelMsg": ""})
|
||
if not cfg.config["interact"]["playSound"]: # 非展板播放
|
||
content = {'Topic': 'Unreal', 'Data': {'Key': 'log', 'Value': ""}}
|
||
wsa_server.get_instance().add_cmd(content)
|
||
if self.deviceConnect is not None:
|
||
self.deviceConnect.close()
|
||
self.deviceConnect = None
|
||
if self.deviceSocket is not None:
|
||
self.deviceSocket.close()
|
||
|