Forklift Accident

The Forklift Accident model detects forklift-related hazardous behaviors in real time from industrial site video.

Feature Summary

  • Pedestrian collision risk detection (risk_1): Forklift operating while a person approaches the work zone within the camera’s field of view
  • Reckless forklift driving detection (risk_2): Person riding on the forklift fork within the camera’s field of view
  • Forklift collision risk detection (risk_3): Forklift operating with obstructed visibility due to overloading

Detection Classes

Class ID Class Name Description
0 normal Normal (no hazard)
1 risk_1 Pedestrian collision risk
2 risk_2 Reckless forklift driving
3 risk_3 Forklift collision risk

Input / Output

Item Details
Input Industrial site image (480×480), work zone boundary info for risk_1
Output Hazard class (0–3), per-class confidence
Recommended Environment TensorRT GPU inference recommended

Demo Library Download Options

Option Description Download
Option A: SDK Inference and post-processing utilities included <python-sdk-download-url>
Option B: Demo Sample Single/batch request samples <rest-sample-download-url>
Option C: Docker Demo Offline demo environment <docker-demo-download-url>

Quick Start (SDK)

C++ 

// TODO: C++ SDK example

License