DFRobot has launched the HUSKYLENS 2, an AI vision YOLO module that is a makers, educator, and embedded developer. This small deep-learning camera is based on 1.6 GHz with a 2-core (K230) and 6 TOPS AI-accelerator providing real-time computer-vision capabilities directly on the device without requiring any cloud processing.
The AI vision YOLO module allows users to use their own vision applications on object detection and recognition by supporting custom-trained YOLO (You Only Look Once) models. It has more than twenty native AI capabilities as well like face recognition, gestures, and object tracking. The HUSKYLENS 2 offers an easy-to-use platform to develop and experiment directly with edge-AI vision systems with the following features.
DFRobot HUSKYLENS 2 specifications:
The HUSKYLENS 2 AI Vision YOLO Module is based on dual-core K230 1.6 GHz processor and a 6 TOPS NPU that provides fast and efficient end devices inference. It has 1 GB of LPDDR4 RAM and 8 GB of eMMC disk drive which enables users to file and execute models without accessing cloud server. The module has a 2 MP GC2093 photo sensor with the capacity of 60 fps, a 2.4-inch IPS touchscreen (640 x 480 pixels) with a user-friendly interface that allows real-time monitoring and setup.
It is also equipped with more than 20 integrated AI models that are useful in the process of recognizing objects, face recognition, pose estimation, line tracking, OCR and barcode or QR code scanning. It also allows users to install their own custom-trained YOLO models, which allows them to be flexible to detection requirements. The camera has a modular shape that can take any type of lens such as the microscope, long-range and night-vision lenses hence it can be used in varied settings and applications.
For connectivity, the HUSKYLENS 2 includes a USB-C port and Gravity-compatible I²C/UART interfaces for seamless integration with controllers such as Arduino, Raspberry Pi, ESP32, and micro:bit. An optional WiFi 6 module enables remote monitoring and data collection. With a compact 70 x 58 x 19 mm footprint and low power consumption of 1.5–3 W, it fits easily into embedded robotics, automation, and portable AI vision systems.
The HUSKYLENS 2 AI Vision YOLO Module features a versatile software stack designed for both beginners and advanced developers. Users can easily operate the device through its built-in touchscreen to access preloaded models, adjust detection thresholds, and fine-tune performance parameters. For advanced workflows, it supports the creation of custom datasets and training of YOLO-based models using frameworks such as ONNX, PyTorch, or TensorFlow, which can then be deployed directly through the onboard model deployment tool. The companion wiki provides detailed wiring diagrams, controller integration examples, and setup guides, along with support for the Model Context Protocol (MCP), enabling the module to transmit structured vision data to large language models for enhanced contextual processing and intelligent edge applications.
The HUSKYLENS 2 AI Vision YOLO Module is available for purchase through DFRobot’s official store at a price of $74.90. Buyers can also choose from several optional accessories, including a WiFi module priced at $7.90 for wireless connectivity and a 30× microscope lens for $11.90 to enhance close-up imaging performance.
Niladri Chowdhury