Yi-Home_Telegram_Bot_Interface

Ditch Xiaomi's cloud and use a Telegram bot instead

Features

• Motion detection

  Works by monitoring a tmp file that is created only when the camera detects a motion

  

• Get a picture from the camera

  Uses the snapshot.sh functionality to get a real time picture

  

• Multi camera support

  Add as many cameras as you please

• The bot works in groups and there's a API limit bypass

  Multiple bot token can be added to round robin the requests (Telegram bots cannot send many messages per second in a group chat)

• Text to speech to all cameras

  Use the speak.sh functionality to convert text to audio. Supports en-US, en-GB, de-DE, es-ES, fr-FR, it-IT

• Text to speech to a single camera

  Same but limited to a single camera

• Play any preconfigured sound to all cameras

  Play any audio from a specified folder (default: sound/). Use ffmpeg to convert it to the needed format

• Play any preconfigured sound to a single camera

  Same but limited to a single camera

  

• Send a voice note and it will play on all cameras

  Send a voice message and ffmpeg will convert it and play simultameusly to all cameras

  

• Alert if a camera goes offline

  No description

• Alert if a camera goes online

  No description

• Function to enable/disable motion notification

  Mute the camera alerts

  

• Function to enable/disable the camera while retaining the audio functionalities

  Disable video and photo functionalities of the camera

  

• Media retention settings

  All media are saved into a folder (default: recording/) and will be deleted in X days. Set 0 to delete immediately

• Multi language

  Currently supports IT and EN languages. You can add more by adding a locale file in config/

  

Note

• No need for Xiaomi's cloud service (The app is needed only for the first setup)

• The cameras don't need to connect to the internet so they will work fine in a isolated vlan

  If you have a UniFi access point you can add a Wi-Fi network with a custom VLAN else you can simply use a dedicated access point

• You'll need to install Yi-Hack firmware on the camera and enable FTP and motion recording

  https://github.com/TheCrypt0/yi-hack-v4 https://github.com/roleoroleo/yi-hack-MStar https://github.com/roleoroleo/yi-hack-Allwinner https://github.com/roleoroleo/yi-hack-Allwinner-v2

• For the text to speech functionality you'll need to install the extra nanotts package

  https://github.com/roleoroleo/yi-hack-utils

• A Raspberry/Docker or LXC Container/VM needs to be connected to the isolated network and to a network with internet access

• There's 2 branches, one with a lot of emojis in the text and one with none

How to use

Edit config/config.py and add your bot api token and the cameras ip address and nickname and then run main.py

Dependencies

ffmpeg and https://github.com/python-telegram-bot/python-telegram-bot

urllib ftplib requests

I'm not sure which one are already packaged with Python 3