Pre-launch · design-partner pilots open for 2027Synthetic defects · edge inspection · NVIDIA stackApply for the pilot

Synvex · synthetic defect data

Generate the defects you cannot wait for

Synvex renders photorealistic defect images from a part's CAD, material and a defect taxonomy, each one labeled to the pixel. Rare defects get as many training examples as common ones.

Specification

Inputs
STEP, USD or OBJ geometry; material specs; defect taxonomy
Outputs
Images with pixel masks, boxes, classes, metadata (COCO and KITTI export)
Rendering
NVIDIA Omniverse Replicator, path-traced on DGX/HGX or cloud GPUs
Randomization
Lighting, pose, camera, background, texture, defect parameters
Capabilities

What Synvex does

  1. 01

    CAD and materials

    Import STEP or USD geometry and material specs, and mark the faces inspection cares about.

    Core
  2. 02

    Procedural defects

    Place scratches, cracks, dents, misalignment, porosity and contamination with controlled size and depth.

    Core
  3. 03

    Domain randomization

    Vary lighting, pose, camera, background and texture so models don't learn the render.

    Core
  4. 04

    Labels for free

    Every frame ships with pixel masks, boxes, classes and full provenance metadata.

    Core
  5. 05

    Dataset versioning

    Diff two runs, pin a dataset to a model, and reproduce any run later.

    Core
  6. 06

    A growing defect library

    Defect classes and realism settings improve with every part and industry (planned).

    Planned
Example output

What Synvex produces

RenderPixel label

Illustrative. Every Synvex frame ships with its label, so there is nothing to annotate by hand.

How it works

Inside Synvex

  1. ST-01

    Scene

    CAD, material and camera intrinsics assemble a scene.

  2. ST-02

    Defects

    The taxonomy places defects within your tolerances.

  3. ST-03

    Render

    Omniverse Replicator renders on DGX or cloud GPUs.

  4. ST-04

    Export

    A versioned dataset goes to Coreon for training.

Defect taxonomy

Example defect-class catalogue

Filter by industry. Classes and label types are examples of the planned starter taxonomy; pilot partners will shape what ships first.

Showing 10 of 10 example defect classes

ScratchSurface
Typical parts
Machined metal, glass, plastics
Labels
Masks, boxes, length and depth
DentSurface
Typical parts
Sheet metal, housings
Labels
Masks, boxes, depth map
Stain and discolorationSurface
Typical parts
Coated and painted parts
Labels
Masks, boxes
CrackStructural
Typical parts
Castings, ceramics, solder joints
Labels
Masks, boxes, branch topology
PorosityStructural
Typical parts
Castings, welds, molded parts
Labels
Masks, pore count
BurrStructural
Typical parts
Machined and stamped edges
Labels
Boxes, edge segments
MisalignmentAssembly
Typical parts
Components, labels, caps
Labels
Boxes, offset and angle
Missing featureAssembly
Typical parts
Fasteners, pins, components
Labels
Boxes, presence flags
Foreign particleContamination
Typical parts
Vials, blisters, food trays
Labels
Masks, particle size
Seal defectContamination
Typical parts
Blister packs, pouches, lids
Labels
Masks, seal-width profile
Specification

Specification

Items marked planned are in design with pilot partners.

ItemDetail
InputsSTEP, USD or OBJ geometry; material specs; defect taxonomy
OutputsImages with pixel masks, boxes, classes, metadata (COCO and KITTI export)
RenderingNVIDIA Omniverse Replicator, path-traced on DGX/HGX or cloud GPUs
RandomizationLighting, pose, camera, background, texture, defect parameters
Rare-defect augmentationWorld and diffusion models for novel defects (planned)
Realism scoringSim-to-real gap estimate per dataset (planned)
Questions

About Synvex

How realistic are the renders?

Realistic enough to train on, which is the test that matters. Coreon measures each dataset against real frames, and Synvex settings are tuned until that gap closes.

Can we add our own defect classes?

Yes. Describe the defect, its size range and where it occurs, and it becomes a class in your taxonomy.

Who owns the generated data?

You do. Generated datasets for your parts belong to you.

Next step

Try Synvex on one SKU

Pilots start with one part, one defect list and accuracy measured on real frames.