Buzzwords

Accuracy measures how well the model can make correct decisions or predictions. It tells us how often the model's decisions match the expected results. It's an important way to evaluate how good a model is at its job.

Machine learning is a type of computer technology that helps machines learn from data without being explicitly programmed. It allows the computer to improve its performance by learning patterns and making predictions based on what it has learned.

Model: Captur has built a special computer system called a machine learning model. This model can make decisions based on the information it receives.

An ontology is a way of organising and describing different things in images. It's like a structured list of categories or types of objects. It helps the model understand the relationships and characteristics of different elements in an image, which helps it analyse and interpret the data more accurately.

Precision is the measure of quality. For example, when we say bad parking how many times did are we right?

Recall is the measure of quantity. For example, when there is an image of bad parking, how often do we spot it?

Signals are how we interpret the image. They’re the kinds of patterns that we try and pick-up from the image. For example, a double-yellow line on the road in london is a signal is used to assess if a bike is in the road.

When we say we're "training" the model, it means we're teaching it to pick up on the signals in the image better.

Visual AI refers to computer systems that can understand and work with visual information, like pictures or videos. It involves creating models and algorithms that can recognise objects, classify images, and generate new visual content.