Unity, one of the leading platforms for building and operating RT3D (real-time 3D) content, has revealed the availability of Unity Computer Vision Datasets. It is launched to reduce the cost of developing computer vision apps and train AI for the Retail, Security, and Manufacturing industries faster. Unity’s platform provides a complete set of software solutions to build, manage and monetize interactive, real-time 2D and 3D content for smartphones, tablets, laptops, consoles, and augmented reality and virtual reality devices.
Synthetic data is essential as it is produced to meet particular needs and conditions that are not ready in existing data. One typical synthetic data use case is for testing a pre-released product when information is not available. Synthetic training data is also a requirement for ML algorithms.
“By creating a synthetic version of datasets that mirror validated privacy rules and accurately reflect real-world data, we enable these groundbreaking datasets to get into the hands of more innovators. Essentially, these datasets empower companies to plan for and simulate scenarios they haven’t yet experienced, with a sizable increase in user data that mimics what they’d find over time in the real world. As a result, we’re seeing smarter indoor environments, such as cashier-less grocery stores, and more as our customers discover new applications.” – Dr. Danny Lange, SPV, AI & ML, Unity.
The Computer Vision Datasets from Unity makes use of a procedure known as “domain randomization” to create several datasets to enhance quality and control bias in apps. The method outputs transformations of how objects of interest are positioned, including camera angles and variances in lighting and as well as the infinite configurations to the Unity space.
Unity’s synthetic datasets also bypass the privacy traps and uncontrolled biases that can arise from methods that usually include images of real places and people from the Internet or captured manually from the real world.
“Synthetic data is revolutionizing the training of machine learning models as it overcomes many of the shortcomings of manually collected and labeled real-world data. Exploring what’s possible and connecting creators with the affordable data they need to make the right decisions continue to drive Unity, no matter the industry. This is why our team will be available to assist customers in ensuring that the datasets produced to meet the right criteria for their needs.” – Dr. Lange.
Follow us on LinkedIn
Read other Articles