Skip to content

code_explanation.md

Latest commit

 

History

History
65 lines (48 loc) · 4.23 KB

code_explanation.md

File metadata and controls

65 lines (48 loc) · 4.23 KB

Everyday Studio

Code Structure:

Here's the code structure, and small explanation of what every module contains and what it does:

📦dev_ws
┣ 📂configs
┃ ┣ 📜env_vars.sh
┃ ┗ 📜predictor_landmarks.dat
┣ 📂export
┃ ┗ 📜every_day.mp4
┣ 📂media
┃ ┣ 📂images
┃ ┗ 📂sound
┃ ┃ ┗ 📜track_1.mp3
┗ 📂src
┃ ┗ 📂everyday_studio
┃ ┃ ┣ 📂utils
┃ ┃ ┃ ┣ 📜errors.py
┃ ┃ ┃ ┗ 📜utils.py
┃ ┃ ┣ 📜capture.py
┃ ┃ ┣ 📜every_day_maker.py
┃ ┃ ┣ 📜face.py
┃ ┃ ┗ 📜file_utils.py

📂 dev_ws: root code folder containing all the scripts, utils, modules for the everyday studio.

📂 configs: Where the config files are located, here you find:

  • 📜 env_vars.sh: bash script with environment variables defined in the runtime for change everyday studio visuals or export settings.
  • 📜 predictor_landmarks.dat: when you run make_video.sh for the first time this file will appear, and are the weights for the face detection model/process.

📂 media: contain all resources and media files for the project:

  • 📂 images: here all images or dataset images should be allocated.
  • 📂 sound: audio files to chose when exporting a video.

📂 src/everyday_studio: folder containing source code of the everyday studio: modules, utils, scripts, functions, and so on. in src folder, also git sub-modules are located.

Code Explanation:

Let's first explore how the everyday studio source code is composed:

┗ 📂everyday_studio
┃ ┣ 📂utils
┃ ┃ ┣ 📜errors.py
┃ ┃ ┗ 📜utils.py
┃ ┣ 📜capture.py
┃ ┣ 📜every_day_maker.py
┃ ┣ 📜face.py
┃ ┗ 📜file_utils.py

that's almost everything. When the bash script make_video.sh is executed, the main() function of every_day_maker.py is compiled, and what is done here is create and instance of the class Studio and execute its method run().

An instance of the class Studiowill load some environment variables, and create 3 instances of other important classes:

  • DataSet loads and handles the dataset specified through environment variable.
  • FaceDetector loads and handles the face detector model and visuals
  • VideoWriter handles the everyday video exportation

In the run() method there's an no end-loop with a while method, here is display the last image by default in the dataset loaded, if any error occurs during the visualization of a image in the dataset or an error in this last one, a black image will be shown. Here. the a OpenCV window waits for an user input, to display and show help, press the key H.

Maybe the most important method in the Studio instance is export_video, when you press the Key C the exportation process is started and then the dataset will be taken from its first element o the last one, and every image from it is passed to the VideoWriter instance to export the video, when the process is completed the idx dataset returns to the previous one, and the Studio instance will display this index.

To see all logs in the pre-visualization or exportation process set the environment variable DEBUG_LEVEL=0.