top of page
Software for Trustworthy Machine Vision AI
Meet Cignal Engine: an environment to safely exercise and test AI vision models for homeland security applications.
Cignal Engine is designed to help scientists, engineers, and testers accelerate time-to-market, optimize performance, and reduce uncertainty for applications that rely on non-visible spectra.
Create digital clones of people, objects, and scenarios without security friction
Generate 2D and 3D labels and annotations automatically for image datasets
Reduce space, cost, and delays with lightweight Cignal Engine Extracting Datasets (CEEDs)
Validation & Test
Verify safe and trustworthy AI using synthetic, multi-modal image streams
The Cignal Engine vision AI platform supports the full AI development lifecycle -
from initial concept, to testing, and final deployment.
Design and create new synthetic objects or CEEDs to test models. Create new weapons, explosives, custom designs, and concealments.
Choose rigid or soft body physics solvers to automatically simulate bag packing, cargo palletizing, or concealment of items on people.
Generate image streams from lightweight CEEDs on-demand, creating massive X-ray, X-ray CT, or millimeter wave datasets at greatly reduced cost.
Cignal Render Core
Scale Cignal Engine nodes horizontally to distribute and share large or critical workloads over multiple GPUs for high performance.
Cignal Engine incorporates a number of innovations to improve the accuracy, performance, and safety of machine vision AI.
Infinite surface, shape, and material perturbation for continuous feature activation.
Create single-snapshot clones of imaging hardware for high-fidelity data generation.
Create massive datasets on-demand with Cignal Engine Extracting Datasets (CEEDs).
Full Spectrum Imagery
X-ray, gamma ray, infrared, and millimeter wave are a few examples of the wavelengths found in advanced security, inspection, screening, and defense use cases.
Traditionally, acquiring training or test data for these spectra is time-consuming and expensive, and it can easily exhaust memory, compute, and storage budgets.
Cignal's innovative approach allows these non-visible datasets to be created, shared, and stored faster, with fewer limitations, and at a greatly reduced cost.
bottom of page