Setting up Website

1. Download streamlit thorugh pip install streamlit
2. Navigate to the directory Website/ where the file main.py is located in the terminal
3. Run streamlit run main.py and open the localhost link to use the website.

Labeling Data

1. Download Label Studio through the steps at Identify Data Annotation Tool
2. Afterwards, create an account once it starts up (can use any credentials, doesn't matter)
3. Click Create Project
4. Name it whatever you want, then go to Data Import and Upload your files

5. Go to labeling setup and click on KeyPoint Labeling

6. Go to the top right where it says visual/code and make sure it's switched to code.

7. Copy paste the following labels into the code box. This will set up the labels for the project

<View>
  <Image name="image" value="$image" zoom="true"/>
  <BrushLabels name="tag" toName="image">
    
    
  <Label value="Good Range of Motion" background="#FFA39E"/><Label value="Bad Range of Motion" background="#D4380D"/><Label value="Straight Back" background="#FFC069"/><Label value="Tilted Back" background="#AD8B00"/><Label value="Symmetrical Squat" background="#D3F261"/><Label value="Asymmetrical Squat" background="#389E0D"/></BrushLabels>
</View>

8. Click Save, you should now see all of your images.

9. To start labeling the images, you can click on Label All Tasks.

10. From there, you can label the image. For example, if someone has good range of motion, click on that label at the bottom of the screen and draw from their heels to their hip. If they have a straight back, click on the straight back label and mark it on the picture. If they have an asymmetrical squat, you can mark that on the photo. Label one feature at a time, then click submit.

11. If you can't accurately discern a label from an image, skip that label, and just add the ones you can. For example, in the above image, you can't accurately discern a symmetrical squat, so you just mark tilted back and good range of motion.

12. After finishing an image, click submit, and it should take you to the next one, where you can continue marking up. At the end, all the images should be marked up, with the home page looking something like this.

13. Repeat this process until you've labeled every image.

13. Once you're done, you can export the data into a json format, which can be downloaded and read later.