Skip to content

[Feature]: Add Google Colab Notebook for Scene Graph Surveillance Reasoning #138

@upasana-2006

Description

@upasana-2006

Problem Statement

Eagle relies on surveillance events and reasoning workflows to analyze activities such as restricted-area intrusion, loitering, crowd formation, and object abandonment. However, contributors and developers currently lack a simple way to generate realistic test data for development, debugging, and experimentation.

Testing often requires manually creating event payloads or obtaining real surveillance datasets, which can slow down development and make reproducing scenarios difficult.

Proposed Solution

Eagle relies on surveillance events and reasoning workflows to analyze activities such as restricted-area intrusion, loitering, crowd formation, and object abandonment. However, contributors and developers currently lack a simple way to generate realistic test data for development, debugging, and experimentation.

Testing often requires manually creating event payloads or obtaining real surveillance datasets, which can slow down development and make reproducing scenarios difficult.

Example output:

{
"person_id": 4,
"event_type": "restricted_zone_intrusion",
"timestamp": "2026-05-31T10:20:11Z",
"location": [120, 85]
}

Suggested location:

utils/synthetic_event_generator.py

Affected Component

LLM Reasoning (services/reasoning/llm.py)

Estimated Difficulty

🟡 Intermediate — Requires understanding of one service

Alternatives Considered

  1. Manual creation of test JSON files
  • Time-consuming
  • Difficult to scale
  • Hard to reproduce complex scenarios
  1. Using real surveillance datasets
  • Privacy concerns
  • Larger storage requirements
  • More difficult setup process
  1. Random event generation without predefined scenarios
  • Easier to implement
  • Produces less realistic and less useful test cases

Additional Context

This feature would help contributors test reasoning pipelines, validate future AI modules, and quickly generate reproducible surveillance scenarios without relying on external datasets.

The generated events can also be used in future notebooks, demonstrations, benchmarking tasks, and automated testing workflows.

Contribution

  • I would like to implement this feature and submit a PR.

Checklist

  • I have searched existing issues and this is not a duplicate.
  • I have read the CONTRIBUTING.md guidelines.

Metadata

Metadata

Assignees

Labels

enhancementNew feature or request

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions