Overview

Momentis is a Python-based project that aims to analyze first-person shooter video game clips for specific keywords (names in kill-feed) and patterns (audio ques: TODO) using OpenCV and other libraries. The main goal of this project is to take several hours of footage from individual clips and trim them down leaving only the relevant parts.

To determine the efficacy of this project, the following show simple metrics comparing the total duration and size of the original footage with the trimmed versions.

Game	Original Duration	Original Size	Result Duration	Result Size	Diff
CS:GO	8 hours	26GB	2 hours 45 mins	8GB	~ 3x Reduction
PUBG	3 hours 45 mins	46GB	30 mins	7GB	~ 7x Reduction
CS2	TODO	TODO	TODO	TODO	TODO
Apex	TODO	TODO	TODO	TODO	TODO

Note: This will re-encoding each video and due to limitations in moviepy api, gpu transcoding is not supported. This results in a long processing time for large video sets. For reference, 8 hours of CS:GO footage took my i7-12700k roughly 4 hours. I suggest running it overnight. Expect high cpu usage.

Getting Started

These instructions will get you up and running on your local machine.

Steps:

1. Install tesseract-ocr and trained data for your language
2. Clone this repository
3. Create a virtual environment
4. Activate your virtual environment
5. Install dependencies
6. Populate keywords.txt with desired words to initialize a subclip (one per line)
7. Run Momentis

Prerequisites

• Python >= 3.12
• python (executable in your PATH)
• git (executable in your PATH)

To ensure python has been installed correctly run the following command:

$ python --version
Python 3.12.6

Setup

As always, using a virtual environment is recommended:

Recommended Steps

# Step 1: Clone this repository
git clone https://github.com/jMujunen/momentis.git
cd momentis
# Step 2: Create a virtual environment (optional but recommended)
python -m venv venv

# Step 3: Activate your virtual environment

# Linux
source venv/bin/activate
# Windows
\venv\Scripts\activate

# Step 4: Install dependencies
pip install -r requirements.txt

Alternative Steps

1. Install tesseract-ocr and trained data for your language
2. Download the zip file by clicking on "Code" -> "Download ZIP"
3. Extract the zip file into any location on your computer
4. Open a shell eg. command prompt, terminal, bash etc.

• Navigate to where you extracted the files using cd /path/to/momentis-master
• Othewise, open a shell in your file explorer by right clicking the Momentis-master and choosing "Open in terminal" or "Open PowerShell window here"

5. Create and activate the virtual environment (optional but recommended)

python -m venv venv
# Linux
source venv/bin/activate
# Windows
\venv\Scripts\activate

5. Install dependencies

pip install -r requirements.txt

Populating keywords.txt

Keywords are used to identify when to start and stop recording a subclip. Several subclips create a single video. You should populate keywords.txt with words that you expect to see in the kill-feed of your game, such as your username. Generally, the OCR (Optical character recognition) has a hard time finding an exact match in certain scenes so its recommended use small 'chunks' instead a of a whole word.

For the clip shown at the bottom, my username is 'MrHoffman'. The following shows how my username gets interpreted by the OCR:

mrhoffman_
mrhoffman_
you finally killed
you knocked out
hrhoffman.
mihoffman_
mrhafiman

Running Momentis

python momentis/momentis.py /path/to/your/videos

This will process the videos in the specified directory and generate output in a subdirectory named "opencv-output"

Showcase

The following shows a glimpse of how it works

Original Duration: 90s Trimmed Duration: 41s

Below is the resulting video