SOP Monitoring Blueprints (Training + Inference)

Table of Contents

• Overview
• When To Use Which Blueprint
• End-to-End Workflow
• Usage
• Sample Data
• License

Overview

Build, train, and deploy a complete SOP (Standard Operating Procedure) monitoring system with two complementary blueprints: - Training Blueprint (sop-training-bp/README.md): prepares a SOP-specific Vision-Language Model (VLM) and Temporal Segmentation Model. - Inference Blueprint (sop-inference-bp/README.md): serves the trained model behind a scalable DeepStream-SOP.

When To Use Which Blueprint

• Use the Training Blueprint if:
• You need to create or refine a SOP-aware model from your own annotated videos.

• You want to programmatically generate QA training data for fine-tuning.

• Use the Inference Blueprint if:

• You already have a trained SOP model (e.g., from the Training Blueprint).
• You want to deploy SOP monitoring as an API and web application.

These blueprints are designed to work together: train a model with the Training Blueprint, then serve it with the Inference Blueprint.

End-to-End Workflow

1. Annotate videos by marking action start/end timestamps (Training).
2. Generate QA pairs (GQA/BCQ/MCQ) from annotations (Training).
3. Fine-tune a VLM (e.g., Cosmos-Reason1) on generated data (Training).
4. Fine-tune a Temporal Segment Model on the annotated data (Training).
5. Deploy the trained model into the Inference Blueprint (Inference).
6. Conduct end-to-end SOP monitoring and analyze SOP compliance (Inference).

Usage

Please refer to: - Training: sop-training-bp/README.md - Inference: sop-inference-bp/README.md

Sample Data

We provide sample data which can be used for testing the SOP Training and Inference BP. The sample data is about installing server fan and power.

License

The software and materials in this repository are governed by the NVIDIA Software and Model Evaluation License

This project will download and install additional third-party open-source software projects. Review the license terms of these open-source projects before use.