Modul Visi AI SingTown SC1

SingTown AI Vision Module makes it easy to integrate AI vision capabilities into your projects. Supports classification and object detection.

Development Method

No software installation or coding required - simply collect images, upload them to the SingTown AI website for annotation, and generate usable models. After connecting to your computer via data cable, you can view real-time video images and running results in the SingTown AI Vision Module's management configuration page in your browser. Models can be directly flashed to the vision module for offline operation.

Integration

SingTown AI Vision Module runs neural network models and sends out results after image recognition. Your system only needs to receive results without requiring image algorithms. The SC1 serial version communicates with your other devices via serial port at 115200 baud rate.

Features

• Supports neural network object detection and classification.
• High performance with built-in 1Tops(int4) 0.5Tops(int8) NPU, offline running YOLOv5s at 15FPS.
• Training is convenient and free, you can get a model in just 10 minutes using SingTown AI training.
• Compact size of 24mm*24mm*16mm for flexible deployment and integration.
• Comes with case, suitable for mass integration.
• Serial port output for results.
• No coding required, no GPU computer needed, offline operation.
• 640×480 color resolution.

• Technical Document: https://docs.singtown.com/ai/zh/latest/index.html