Ru En
Chatbot presentation from George Fomitchev

Download Endurance chatbot platform ver. 4.0


Endurance offers an open-source platform for creation of a robot-companion for Android and Windows 7 / 8 / 10. htmlimage The platform consists of the following modules:
  • Speech recognition from Ok Google.
  • Image (face) recognition on the OpenCV basis.
  • Microsoft SAPI.
  • Botlibre (a platform for ChatBot creation and setting) with the possibility of integration with Microsoft Botframework.
  • platforma The main Botlibre settings Users - Configure who can access, and administer your bot. Avatar - Configure your bot's appearance. Choose an animated avatar, or create your own. Voice - Configure your bot's language and voice. Learning & Settings - Configure your bot's learning ability and other settings. Training & Chat Logs - Train your bot's responses, greetings, and default responses. View and your bot's conversations. Import and export chat logs to text or AIML files. Twitter - Allow your bot to manage a Twitter account and interact with other Twitter users. Facebook - Allow your bot to manage a Facebook account or page and interact with other Facebook users. Telegram - Allow bot to manage a Telegram channel or chat on Telegram. Slack - Allow your bot to send, receive, and reply to Slack messages. Email - Allow your bot to manage an email account and answer emails. SMS - Allow your bot to send, receive, and reply to SMS messages. Google - Allow your bot to connect to Google services such as Google Calendar. IRC - Allow your bot to chat with others on an IRC chat channel. Web - Import data from the Freebase, Wiktionary, or other websites. Knowledge - Browse your bot's knowledge database. Scripts - Add, create, edit, import, and export Self or AIML scripting programs. Log - View the bot's log for errors and debugging info. This software complex consists of 4 modules enabling to create an embedded robot-companion (ChatBot) for OS Windows 7/ 8/10.

    Advantages

    The key feature of the Endurance solution is as follows: besides the standard environments, such as: Web, Slack, Telegram, Twitter, Facebook, you can also use the embedded version of ChatBot offline. In addition, the embedded solution of Endurance includes a built-in face recognition module, that is, a ChatBot sees who he talks with and is able to identify the interlocutor by face.

    How does it work?

  • You run the app. for Windows 7 / 8 / 10.
  • The program identifies the user (first start) and asks to enter his name or recognizes the user by his face (repeated start).
  • The speech recognition program (from Ok Google) starts running. Depending on the settings, the pronounced phrase is heard through and analyzed during 30 seconds as a whole, or is heard through and analyzed piece by piece.
  • The phrase is recognized and sent to the Chatbot on Botlibre.
  • Botlibre sends back the response to the program.
  • The program synthesizes the voice using Microsoft SAPI.
  • Further on, if the conversation keeps going, the process is repeated. The conversation can last indefinitely long.
  • htmlimage (1)

    Who and what for needs it

    First and foremost this solution may be used for training by those who dream to create one’s own real robot-companion, including university and high school students. Creating one’s own ChatBot, you acquire knowledge and practice in developing embedded solutions for Windows 7 / 8 / 10, and work experience with:
  • С++,
  • AIML Markup language,
  • OpenCV,
  • JSON requests.
  • platforma2 Any enthusiast can create one’s own ChatBot for Web, Telegram, Facebook, Slack, Twitter, Gmail, etc. One can use a robot-companion for scientific and practical purposes. As:
  • A platform for integration with physical (wheeled, for instance) robots like a robot-waiter, one mile delivery, a telepresence robot, etc.
  • A robot-teacher of foreign languages.
  • A robot-receptionist аt the reception desk.
  • A robot-companion for senior people.
  • htmlimage (1) What we offer
  • A free Endurance Chatbot Platform ver. 4.0.
  • Support.
  • Tuition.
  • Development of a ready ChatBot for a particular order or a business case.
  • Endurance Chatbot Platform ver. 4.0 settings In the source code there are a few open settings: bot_app_id=1156124978941390219 // ChatBot’s name for connection bot_instance_id=14142865 // profile name bot_address=www.Botlibre.com // connection address to Botlibre server recogn_lang=en-US // speech synthesis language speech_lang=409 // speech recognition language google_key=AIzaSyBT3lf4Z3mM5vyEEGJTWSM_T77sKpdDrUk // connection key to Ok Google for speech recognition sound_record_level=10000 // relative level of the audio signal for automatic speech recognition start pause_record_msec=2000 // duration of the last pause before a phrase is sent to Ok Google to be recognized speech_speed=0 // relative speech synthesis speed Other settings can be additionally done.

    About our solutions

    The platform is developed under the Creative common license Attribution-NonCommercial-ShareAlike 3.0 htmlimage (2) The platform is offered «as is», and the developers are not responsible for the disruption of any features that are provided by third-party vendors, such as: Google, OpenCV, Microsoft, Botlibre.

    About the developers team

    George Fomitchev, Endurance founder
  • Speaker at MoNage, Boston 21 Sep, 2016.
  • Finalist of Pitchfire RoboBusiness, San Jose, 28-29 Sep, 2016.
  • Speaker at the annual conference IEEE Connecticut Conference on Industrial Electronics, Technology & Automation (CT-IETA 2016), Bridgeport, 14 - 15, 2016 http://ctieta.org/keynote.html
  • Andrey Azovsky , software developer with 15 year experience C / C++ / JSON. Anatoly Preobrazhensky, ChatBot developer in the Botlibre environment. Serguey Mironov, partner with 5 year experience in commercial sales. Philipp Khokhlov, designer and test person.

    You are welcome to join the project

    The project is entirely open source. We invite to join the project university developers’ teams, individual developers and enthusiasts. htmlimage (2) The project opens great opportunities in learning the technology of ChatBot creation and acquiring work experience in this technology. The project tasks at hand: Develop an application for iOS. Create a robot-reception (visitors registration). http://www.slideshare.net/GeorgeFomitchev/selfiebot-application-for-retail-shops-or-stores http://www.slideshare.net/GeorgeFomitchev/selfiebot-receptionist Create a more user-friendly interface (UX/UI). Create machine learning. Ensure conversion of unstructured texts into structured ones suitable for ChatBots. htmlimage (3) How the robot-companion works The robot-companion hears the speech addressed to it and gives answers based on the programmed variants of answers. In case of their lack it uses third-party search tools, Google, for instance. When the answer is found, the robot gives it to its companion.

    How it is configured

    The system of ChatBot learning is very simple:
  • Templates are uploaded: question / answer.
  • Unstructured texts are uploaded as freebase for independent ChatBot learning.
  • When the answer does not correspond to the question the administrator or the user can offer an improved version of the answer (depending on the learning settings).
  • Usage of XML / JSON requests to get the data to be imported to the ChatBot.
  • How it works

    1. ChatBot creation on Botlibre. 2. ChatBot learning. 3. One’s Google key registration to use «Ok Google» speech recognition. 4. Make necessary changes in the *.cfg. file 5. Program start.

    Educational Program

    The Open-Source Speech and Image Recognition and Speech Synthesis Platform for Educational Needs An Open Source fully 3D printed Platform for Creation of One’s Own Robot-Companion

    Educational Program — «Robotics: Design Engineering and Programming»