HYPER REALISTIC SIMULATOR
BUILT FOR MACHINE LEARNING
Train, test, validate
Unleash the limitations of the physical world with a state-of-the-art simulation platform. SynCity, developed by CVEdia, constructs synthetic 3D real or fictitious environments and provides sensor simulations robust enough to train, validate, and verify artificial intelligence algorithms for land, aerial, and marine autonomous applications.
Providing an infinite amount of environments, SynCity can be used as a developmental and operational test and evaluation platform, assisting with certifying sensor and/or autonomous systems at any stage of development.
ADVANCED SENSOR SIMULATIONS
IN REAL TIME
Identify, detect, track
Generating ground truth sensor data for LiDAR, radar, near and far infrared, thermal and RGB cameras, and IMU/GPS devices, we provide a robust representation of the world. Sensor data simulated on SynCity is indistinguishable to data collected externally.
Simulating off-the-shelf and custom sensor models, we mimic degradation in various operating conditions. SynCity sensor models are designed as plug-ins enabling customization. SynCity provides a combined software-in-the-loop & hardware-in-the-loop solution allowing for a flexible setup.
CONFIGURABLE SCENE ELEMENTS
Flexible API, customize, control
SynCity provides you with an environment where you are able to independently control nearly every possible parameter. Be the master of your scenes and choose from hundreds of thousands of high quality 3D objects. On SynCity, you can simulate all kinds of events such as near misses and crashes, control the behavior of assets, select night/day cycles, and various kinds of weather conditions to test how your application and sensors perform.
SynCity API and GUI systems provide you with an unparalleled degree of customization and flexibility.
Simulation, big data management, testing
The minute you begin running your simulations you are generating billions of images. SynCity provides a turnkey solution with its big data management tool allowing you to easily and quickly filter content, annotate to benchmark against real-world data, and augment images for more entropic datasets. Shake up your models to ensure generalization for quicker algorithm development and validation.