Model Library

Choose the models your runtime uses.

Hearthlight’s Model Zoo feeds the mounted inventory and the stage bindings used at runtime. It exists so operators can change model selection without turning that choice into brittle one-off config edits.

Stage Coverage

detect, track and identify behavior

The current repository describes the Model Zoo as the source for detector, tracker, heuristic filter, and anomaly detection options.

Stage 1

Detector

The detector model decides which raw classes are visible for each source and which class metadata can feed alert-rule options.

Stage 2

Tracker

Trackers maintain continuity between frames so the runtime can pass stable motion and entity context downstream.

Stage 3

Heuristic Filter

Filtering stages refine or gate runtime behavior before later steps consume those signals.

Stage 4

Anomaly Detection

Local and API-backed anomaly stages can surface higher-level activity and object events after the earlier runtime stages finish their work.

Choose Models

Set a default, then tune each camera.

Pick the model you want Hearthlight to use by default. If one camera needs a different detector, tracker, or behavior model, override that camera without changing the rest of the system.

Add Models

Start with the included catalog, then add your own.

Use the built-in options first. When you need more coverage, add models from the model zoo or mount a custom model so it appears as a selectable option.

Operator Experience

The library exists for readable choices, not raw registry spelunking.

Hearthlight keeps stable internal keys for storage and automation while presenting readable display names in the operator-facing surfaces.

Inspect the current inventory

Use the CLI to list templates and model options before startup.

hearthlight list-models

Mount models during onboarding

The onboarding flow can seed the mounted inventory with default models or specific registry keys through repeated `--mount-model` flags.

Contribute

Expand the model surface responsibly.

Contributions can target the runtime’s registry-backed model flow or the dedicated model-zoo repository, depending on whether you are adding catalog coverage or model assets.