A high fidelity simulator for Machine Learning projects


Used to develop autonomous applications, ADAS, and smart sensors, CVEDIA’s SynCity is a powerful simulation platform used to generate data for neural network training and validation. Use SynCity as a custom pocket laboratory to generate highly entropic scenes, conditions, and metadata. Enable real-time Hardware-In-the-Loop (HWIL), Human-In-the-Loop (HITL) or Software-In-the-Loop (SIL) simulations even with complex sensor configurations.




CVEDIA creates fully entropic environments

Three levels of environments, from low definition meshes, no objects, low entropy and no post-pro all the way up to high definition, maximum amount of objects, high entropy and complex post-pro effects.


Carousel imageCarousel imageCarousel image



  • Velocity.

  • Points of interest (IAO/POI).

  • Angular Velocity (3D).

  • 2D and 3D bounding Boxes.

  • World position.

  • GPS position

  • Object state.

  • Optical flow.

  • Object metadata.

  • Pixel segmentation.

  • Object arbitrary metadata.

  • Semantic segmentation.

  • Point clouds.

  • Polygon annotation.

Retrieve metadata and annotations

Our simulator retrieves extensive ground truth in real time from its simulations and generates many annotation types.





Want to see our AI in action?

Book a demo now →




Empirical sensor modelling and fusion

SynCity is the only simulation platform engineered at the protocol communication level. Design simulations with a number of sensor options including LiDAR, radar, infrared, thermal, ultrasonic, RGB, RGB-D cameras, and IMU / GPS devices to synthesize photo-realistic environments. Employ and couple multiple sensors in your training set in real time, and even adjust sensor placement conditions.



BLEND FIELD AND SIMULATED DATA

Use existing field data as a push off point for simulated dataset creation.


GENERATE SENSOR NOISE PROFILES

A team of AI experts to ensure perfection in your deployment, at your disposal at all times.


SIMULATE REALISTIC SENSOR CONDITIONS

Custom algorithms to include unique classification parameters and extra detection classes in just weeks, with no data collection required.




Fully configurable parameters

SynCity provides a custom environment to control user defined perimeters. Build computer vision scenarios with thousands of 3D objects, simulate events including near misses and crashes, and test sensor performance in various lighting, weather, and sensor conditions. Control the behavior of your assets and environments with Python and C++ APIs and a GUI system.


ACCOUNT FOR EDGE CASE CONDITIONS

Test AI system performance in extreme temperatures, darkness, or with glare or water on the lens.


TEST DANGEROUS OR COSTLY EVENTS

Use SynCity to ensure your machine learning system is both safe and within budget.


MODIFY SPAWNING BEHAVIOUR

Adjust object placement and behaviour.



CVEDIA | API's, 3rd party integration, and training


API's, 3rd party integration, and training

SynCity projects are created to be custom and customizable for each client. SynCity integrates with common third party applications and data management tools. Use C++ or Python API to develop a custom workflow for your team, or work with CVEDIA’s in-house team to train and validate your system.