AI is a branch of computer science in which intelligent behaviour is simulated by computer systems. It is often also referred to as a computer system itself that is able to perform and automate tasks such as visual interpretation or image analysis, decision-making, language processing, etc.
At Aiforia we refer to our deep learning artificial intelligence algorithms as AI models. These models are a set of rules or instructions with which the image analysis task is performed. For example Aiforia’s neural network’s can be trained to form an AI model to quantify neurons (brain cells).
Artificial neural networks (ANNs)
ANNs are what form and drive deep learning. They are computing systems designed to find patterns that are too complex to be manually taught to machines to recognize. To find out more about neural networks read our articles: ANN Part 1 and ANN Part 2
Convolutional neural networks (CNNs)
CNNs are an incredibly powerful type of artificial neural network particularly adept at image recognition, enabling AI to reach new heights in image analysis, far surpassing human capability in most cases. CNNs are named after convolutions, a type of mathematical operation which they use to assess input data to extract information.
So how do CNNs analyze images? CNNs decipher sensory data, such as images or sound, through a sort of machine perception by recognizing patterns as numerical. This basically means that sensory information is translated into numbers. To find out more about how convolutional neural networks perform image analysis read our articles: ANN Part 1 and ANN Part 2
Deep learning (DL)
DL is a subset of machine learning (ML), which is a subset of AI. Deep learning is the next generation of ML as it is more complex in its architecture and therefore can solve more complex problems, such as in image analysis. What brings the deep to deep learning is that it is formed of more layers of neural networks than ML. To find out more about the differences between DL and ML, read our article here: DL vs ML
Graphics processing unit (GPU)
GPUs are computer hardware that render images to for example computer monitors. They convert data into a signal that a monitor can understand. They are what have allowed AI to improve its performance and capabilities vastly.
This is the reality, or the expected result, you want your AI model to predict. It is a concept used in supervised learning to make sure your AI model is looking for what you want it to look for. Check out our ground truth video from our guide on Training AI.
Machine learning (ML)
ML is a subfield of AI that enables computers to learn from examples, without explicitly being programmed to do so. Machine learning can be formed in many different ways including from neural networks; these networks are an arrangement of three or four layers of connected computational units.
This is a learning method for AI models sitting between supervised and unsupervised learning. It uses a mix of labeled and unlabelled data, which can be useful when the labelling of the data might be very consuming, resource intensive, or just difficult to extract relevant features from that data.
A method by which AI models are taught. Aiforia’s AI is in most cases trained with the help of supervised learning, in which the neural networks learn with a labeled dataset.
A method by which AI models are taught. Information is gathered by the neural networks of the AI model from an unlabelled dataset by extracting features and patterns on its own. Essentially there is no defined ground-truth and therefore it is a useful method more for generating hypotheses rather than testing them.