5. Synthetic Data Generation (SDG) User Flow
The Synthetic Data Generation (SDG) flow enables you to create new scenarios based on real-world data. It begins by importing a ClipGT file into Foretify Manager, where you can compare the raw data with the denoised run. You can then open an editor to select which actors from the original clip to include, modify their behavior, and add new synthetic actors with custom trajectories. Finally, you can simulate the new scenario and export the results back to a ClipGT file.
| Step | Description |
|---|---|
| Import Clip | Import ClipGT, denoise, and ingest into Foretify Manager. |
| Evaluate Clip | View the ingested run and compare it with the noisy raw data. |
| Edit the Scenario | Select your editing mode. |
| Edit Real-world Actors | Select and modify real-world actors. |
| Add Synthetic Actors | Add and edit new synthetic actors. |
| Playback | Validate selections with a quick replay in the editor. |
| Preview | Run a fast Model SSP simulation to visualize realistic behavior. |
| Simulate | Execute full simulation. |
| Launch | Save the scenario as a test and run it. |
| Export | Save the created scenario as a new ClipGT file. |