By default, TensorFlow Lite Model Maker uses a model called , which is designed to recognize 1000 types of image.

By default, TensorFlow Lite Model Maker uses a model called MobileNetV2, which is designed to recognize 1000 types of images from the ImageNet dataset. MobileNetV2 is streamlined for mobile and embedded vision applications due to its efficiency and lower computational requirements, making it suitable for real-time inference on devices with limited resources. Its architecture is lightweight yet powerful, allowing it to perform well in image classification tasks. The integration of TensorFlow Lite with MobileNetV2 further enhances its usability for deploying machine learning models on mobile and edge devices, enabling developers to implement robust image recognition functionalities in their applications.