vairasza / dhcv

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

README

Purpose

bundles all CV methods for the digital humanities project Distant Viewing of Youtube Game Walkthroughs

Installation

  • create virtual env with Python 3.10.
  • activate virtual env for project path.
  • use python -m pip install -r requirements.txt to install dependencies.
  • check if all dependencies are installed by importing dependencies in all_cv.ipynb.
  • detectron2 must be installed manually. script is in all_cv.ipynb.
  • following error can be ignored: ERROR: Cannot install -r requirements.txt (line 40), google-auth==1.4.2 and tensorboard==2.11.2 because these package versions have conflicting dependencies.

Changing Dependencies

  • pip freeze > requirements.txt

google drive output path

G:\My Drive\digital_humanities\Ergebnisse

google drive input path

G:\My Drive\digital_humanities\Walkthroughs

Running

  • configurate in config.json which methods and games you want to run
  • available methods and games listed here

methods

  • Location
  • AverageColor
  • ObjectDetection
  • Emotion
  • AgeGender
  • Brightness
  • DominantColor

games

  • Die_Siedler3
  • Diablo2
  • Elden_Ring
  • Half_Life2
  • Little_Nightmares
  • NFSMW
  • Starcraft2
  • Super_Mario_World

Running (2)

  • set inputRoot to folder where your raw images are located. can be mounted google drive.
  • inputRoot requires a folder for each game.
  • set outputRoot to folder where your generated csv, graphs and images are saved. can be mounted google drive.
  • outputRoot require a folder for each method which further requires a folder for each game.
  • the naming of methods and games must be exactly like mentioned above.
  • if you do not want to run object detection, uncomment detectron2 related imports
  • code blocks that have #required in the first line, must be loaded before #run block
  • code blocks that have #optional in the first line, must be loaded if there methods are configured in config.json
  • Object Detection requires a GPU

Quellenangaben

About

License:MIT License


Languages

Language:Jupyter Notebook 99.6%Language:Python 0.4%