Guides

Everything you need to use react-native-pose-detection in your app.

Building on the library? You're in the right place. Working on the library? See docs/.

Start here

1. Getting started	Install → live skeleton in five minutes
2. Installation	Expo, bare RN, EAS, release builds
3. Camera control	Switching, pausing, the three toggles
4. Triggers	Business logic that runs natively
5. Data delivery	Getting landmarks out without paying for them

Going further

Photos and video files	Landmarks from a file, and painted copies of one
What you can build	Trigger syntax, feasibility, honest limits
Performance	Profiles, the governor, thermal ladder, app size
Troubleshooting	When something doesn't work, and the log channel

API reference

<PoseCamera> props	Every prop and its default
Events	Every callback, payload, and error code
Ref methods	switchCamera, snapshot, getState, …
Functions	detectOnImage, exportPose, setLogLevel, every other export
Types	PoseFrame, JointName, wire format
Trigger schema	Conditions, emit modes, validation
Camera permission	useCameraPermission, and why blocked is not denied
Config plugin	app.json options
CLI	fetch-model, doctor, clear-cache

Getting the best performance

Four decisions are yours and they dominate everything else:

Decision	Impact	Where
Keep data.mode: 'off' and use triggers	60 bridge crossings/sec → ~1 per event	data delivery
Pick the right model	lite adds ~5.5 MB, heavy ~29.2 MB	performance
Use select to trim the payload	three joints is 48 bytes a frame, not 528	data delivery
Set active={isFocused}	camera off when the screen isn't visible	camera control

Everything else, the delegate, the frame rate, the resolutions and the thermal response, is measured and chosen for you, and every one of those choices can still be overridden per axis.