Microsoft Azure AI Fundamentals (AI-900) Practice Exam

What service can you use to train an object detection model with your own images?

1. Object detection

2. Custom vision

3. Facial recognition

4. Optical character recognition (OCR)

The correct answer is: Custom vision

The service to train an object detection model with your own images is Custom Vision. Custom Vision is a part of Azure's Cognitive Services that allows users to build, customize, and deploy machine learning models for image classification and object detection. This service provides an easy-to-use interface where you can upload your images, tag them according to the objects you want to detect, and then train a model based on these labeled images. Once the model is trained, it can be used to identify and classify objects within new images, offering flexibility and adaptability for various applications. Custom Vision also allows for ongoing improvements by retraining the model with new data, ensuring that the model remains accurate over time. The other services listed, while applicable to different use cases, do not cater specifically to the training of object detection models with custom datasets. For example, facial recognition primarily deals with the identification of faces in images, whereas optical character recognition (OCR) is focused on reading text from images. Similarly, the term "object detection" refers to the broader concept rather than a specific service to build and train models.