Image recognition is very widely used in machine learning. There are many different approaches and solutions to it, but none of them fitted our needs. We needed a completely local solution running on a tiny computer to deliver the recognition results to a cloud service. This article describes our approach to building an object recognition solution with TensorFlow.
It is simply impossible not to notice how quickly the total volume of metrics collected grows. Not only the frequency with which automatic systems collect data and the throughput of data warehouses has increased, but also the set of metrics that we can use. This trend is most clearly expressed in IoT, but other industries can also boast of a huge set of data sources, public or available by a special subscription.
Increasing the amount of data creates new challenges for analysts and professionals working …
The client is a EU-based startup company. They use an automated approach to monitor their installations for utility companies. Their main objectives are managing assets and improving maintenance processes using fast drones.
The client collects images of the objects maintained using drones. The images are collected at high speed, and not all the images contain useful information. This increases the need for a human workforce to filter the images before processing.
The business challenge was to improve the …
In recent years, there has been significant progress in the field of machine learning. Much of this progress can be attributed to the increasing usage of graphics processing units (GPUs) to accelerate the training of machine learning models. In particular, the extra computational power has lead to the popularization of Deep Learning – the use of complex, multi-level neural networks to create models, capable of feature detection from large amounts of unlabeled training data.
GPUs are so well-suited to deep learning because the type of …