Your one-stop solution that offers advanced real-time object detection, automated model training, and extensive admin capabilities to support policing efforts efficiently.
- Check it out here!
-
🕵️Advanced Object Detection
Utilizing fine-tuned computer vision models like YOLOv11, Llama 3.2, Florence, and SAM2, HTX Sentinel can detect objects and provide real-time notifications for quick response. -
🤖Automated Training Pipeline
Our platform includes a fully automated pipeline for computer vision (CV) model training, covering all stages from segmentation, labeling, and training to deployment. -
🧠Comprehensive Admin Capabilities
The admin dashboard offers a detailed view of emergency calls, officer deployment, open cases, and overall incident statistics. The data visualization is interactive, allowing for quick insights into real-time events and recent activities. -
🗺️Live Patrol Map
A live patrol map simulates the real-time locations of officers and displays active incident locations, giving admins a clear operational overview. -
🏋️Streamlined Training Workflow
Quickly train CV models through an intuitive process. Choose segments, provide descriptions, and start the training process with a single click. The platform uses transfer learning to significantly reduce training time while maintaining high accuracy by freezing pre-trained layers and training only the final layer.
HTX Sentinel is built with state-of-the-art technologies to ensure reliable, high-performance service:
- YOLOv11 for object detection
- Llama 3.2 for enhanced AI processing
- Florence for advanced image processing
- SAM2 for automated segmentation and masking
- DeepSort for object reidentification
To deploy HTX Sentinel on your system, refer to the following sections for setup, configuration, and usage.
- Installation Instructions: Follow detailed installation and configuration steps.
- Admin Setup: Set up user roles and permissions, configure settings, and deploy the patrol map.
- Using the Platform: Explore the training and model management features to fully utilize the platform’s capabilities.
