Blackbeard - Team Jaqen

Team Members:

• Nicolas Morant aka DevGlitch
• Huayu (Jack) Tsu aka codejacktsu

Description

Team Jaqen wanted to build Blackbeard, a world class blackjack AI, that trains your ability to play blackjack from a wearable device in real time. The challenges Blackbeard is facing are as follow: object detection, object classification, mathematical probabilities, hand motion recognition, and more.

Disclaimer: The device is intended for academic use or training purposes only. Card counting is not illegal under Federal, State, or Local law.

Project Report

https://github.com/DevGlitch/jaqen/tree/master/project_report/

Project Presentation & Demo

https://youtu.be/R2JZ_oEcPRQ

Getting Started

Follow the below instructions in order to run Blackbeard on your machine.

Installation

1. Clone the repo

   git clone https://github.com/DevGlitch/jaqen.git

2. Install any missing packages.
   • We recommend using a conda environment with Python 3.7.
   • Install all the missing packages on your machine
   • On the Raspeberry Pi:
     • Setting up a RTSP feed is needed
       • We used v42lrtspserver with h624
       • Make sure to adjust the settings of the v42lrtspserver
     • Put the mqtt_client.py file on your Pi

Running

1. Launch on the Raspberry Pi:

   v42lrtspserver

   python3 mqtt_client.py

2. Start Blackbeard on your Desktop:

   python __main__.py

MQTT Message

Count: advantage of player vs house. Higher the count, the more advantage the player has

Opt Action: Optimal action

Bet: Suggested next round's betting amount

License

Distributed under the MIT License. See LICENSE for more information.

Contact

Nicolas Morant - Personal Website & Github

Huayu (Jack) Tsu - Github

Project Link: https://github.com/DevGlitch/jaqen