Commonly Asked Questions:
How much data is required for an AI training set?
We start from 1,000 instances per class, but we work closely with clients’ Machine Learning (ML) engineers during training, validation and testing data to achieve high accuracy of ML models.
How long does it take to annotate AI training set?
Typically, we can turn around the annotation results within 2-3 days, as well as set up the workflow that fits your active learning process.
What is the process of custom datasets development?
We start with collecting and optimizing imagery and map tiles using web data sources, defining the classes, and training Machine Learning classification or object detection models.
How to ensure accuracy of deep learning solution
for my dataset?
We develop Computer Vision model through an iterative process and large validated datasets using supervised learning, followed by additional human-led validation.