Street Sign Sense

🚀 What is Street Sign Sense?

Street Sign Sense is an object detection model designed for the real-time identification and localization of road signs.

This demo demonstrates the power of modern object detection models in analyzing images and video streams to extract information from the environment.

⚙️ How It Works

The system uses a state-of-the-art model from the YOLO12 (You Only Look Once) family, an algorithm known for its exceptional balance between accuracy and inference speed.

1. Model Selection

You can choose from different variants, such as YOLO12n (Nano). The "Nano" model is optimized for maximum speed (high FPS).

2. Input

The model can analyze various sources: a static image (Upload File), an instant snapshot (Take Photos), or a real-time video stream (Webcam).

3. Parameter Adjustment

Confidence: Defines the certainty threshold (from 0 to 1) required for the model to label an object. Lower values increase the number of detections but also the risk of false positives.
IOU Threshold: The Intersection over Union (IOU) threshold is used by Non-Maximum Suppression (NMS) to manage duplicate detections, eliminating bounding boxes that overlap beyond the set threshold.

4. Output

The system highlights detected signs and measures performance in terms of FPS (Frames Per Second) and Inference (time in milliseconds to analyze a single frame).