Terran timelines

Creating face-recognition timelines on videos has never been so easy! Using the power of Terran we can easily build these timelines.

Installation

This Streamlit component requires the following packages for working properly:

# Install dependencies
pip install --upgrade streamlit terran youtube-dl

# Install the component
pip install streamlit-terran-timeline

Usage

You can generate a timeline from any kind of video using the generate_timeline function and then using the terran_timeline Streamlit component like this:

import streamlit as st
from streamlit_terran_timeline import generate_timeline, terran_timeline

# Generate the timeline information
timeline = generate_timeline("https://www.youtube.com/watch?v=dQw4w9WgXcQ")

#
# Display the timeline. If the users click, you'll get the exact second of
# the part of the timeline video. By default, it returns 0.
#
start_time = terran_timeline(timeline)

st.write(f"User clicked on second {start_time}")

You can also check out more examples in the examples folder.

Development process

1. First, switch the _RELEASE variable from streamlit_terran_timeline/__init__.py to False.
2. Then, start a development server at streamlit_terran_timeline/frontend by running npm install and then npm run start
3. Also, you'll need to install the package internally like pip install -e .
4. Finally, run streamlit on and use the component! For example, you can run streamlit run streamlit_terran_timeline/examples/youtube.py

What's Terran?

Terran is human-perception library made by Pento 🚀

With Terran, making this demo was super easy! You can take a look at the generate_timeline function to understand how Terran modules works with videos, face-recognition, and face-detection.