Privacy

Privacy policy

Applies to the iPhone and iPad version of Sign Motion with 2D avatars, Camera only recording, video editing and optional audio recording
Effective: October 4, 2026 / Last updated: October 9, 2026

Face, body and hand tracking runs on your device.

Face images and coordinates used for tracking are not sent to the developer’s or SDK providers’ servers. Videos you choose to record contain drawn facial outlines and movements. Camera only recordings also contain actual faces and surroundings.

1. Developer and contact

Developer: KONDO FUMIKAZU
kondofumikazu@icloud.com

Please use this address for questions about the handling of personal information.

2. Camera, face, body and hand data

2.1 Face data accessed and generated

In animal and skeleton modes, after camera permission is granted, the app processes camera images containing faces on the device and estimates facial landmark coordinates for the face outline, nose, eyebrows, eyes, eyelids, iris area and lips. The recognition SDK may internally estimate relative depth. The app discards z and world-coordinate outputs and retains only normalized x and y coordinates. It also estimates facial expression coefficients for cheek puff, smiles, blinking and mouth movements on the device, using them to render skeletal lines and 2D animal faces. Blinking and mouth opening are displayed as changes in these positions. Body and hand positions are also estimated. Eyebrow, gaze, lip, jaw and nose expression coefficients are used, and separate cheek shape is estimated against a neutral front-facing baseline. Teeth and tongue display uses transient color samples from a limited inner-mouth region. It may confuse tongue with gums and is not precise tongue tracking.

The app estimates head position, size and tilt in the camera image and applies these values to a generic 2D animal face. Values and observation times are handled temporarily in memory. Image positions are estimates. Depth images and camera calibration data are not acquired. The app does not create personal face scans or face authentication templates and does not use ARKit face tracking.

2.2 Uses of face data

In animal and skeleton modes, face data is used only to display skeletal lines and animate a 2D animal face on screen and to include that rendering in recordings started by the user. It is not used for authentication, identifying people, profiling, advertising, marketing, analytics, AI model training or sale. Camera only uses camera images for preview and user-started camera video recordings, without face, body or hand tracking. No other use of face data is planned.

The app runs the bundled MediaPipe Tasks Vision SDK and Apple Vision locally. Face images and landmarks are processed within the user’s device. The app does not transmit them to Google, Apple, the developer or other third-party servers. The developer does not receive face data or provide, share or sell it to advertising platforms, analytics providers, data brokers or other third parties. User-directed saving and sharing of rendered recordings are described in Section 3.

3. Recordings, storage locations and user-directed sharing

The app creates an MP4 only when you start recording. In animal and skeleton modes it records the 2D avatar on a charcoal gray background. It includes 2D animal faces, blinks, mouth opening and body and finger movements, so it contains information derived from face data. These modes contain no camera footage. Selecting the camera button next to the animal buttons switches to Camera only, which records actual faces, bodies, hands and surroundings without skeleton or animal overlays. The app explains before recording that actual footage is included. Audio recording is off by default. If you enable it in Settings and start recording, microphone audio is included. Face images acquired for tracking, numerical landmark arrays are not saved as separate files or databases.

iOS recordings are stored in Documents/Recordings within the app’s private local container. There is no developer-operated server or app-operated cloud storage for face data or recordings.

Video editing imports only the video you choose in the system photo picker and copies it to Documents/ImportedVideos. Any face footage and audio in that selected video remain in the copy. The app does not run face recognition on imported videos or convert them to avatars. Caption and trimming edits create local edited copies and leave the original unchanged. The app does not request permission to browse the full photo library.

On iOS, a copy goes to Photos or the selected sharing destination only when you choose Save to Photos or Share. The app does not automatically transmit recordings.

Copies may leave the device through iCloud Photos, file synchronization, operating-system backups or sharing services according to your settings. Those services determine their storage locations, retention and deletion practices. The app does not implement its own automatic synchronization.

4. Permissions and withdrawing consent

Camera: Used to preview and record camera footage and, in animal and skeleton modes, recognize face, body and hand movements locally. The app requests system permission before access. You can revoke permission in system camera settings.
Add to Photos (iOS): Requested when saving a recording to Photos. This does not request access to read existing photos.
Microphone: Permission is requested when starting a recording with audio enabled. Microphone audio is used in that recording. You can turn audio recording off in Settings.

To stop camera use, put the app in the background. Revoking permission does not delete existing recordings; see Section 5.

The app does not use contacts, location or advertising identifiers. It has no advertising SDK, analytics SDK, developer-operated automatic crash reporting or account registration.

5. Face data retention and deletion

5.1 Images and coordinates used for tracking

Face images, landmarks, expression coefficients, oral color samples, image-plane head tilt and rendered 2D images used for tracking and display are processed temporarily only in volatile memory on the device. Frames are replaced or released as processing advances. Latest results and SDK processing state may remain in memory while the app process runs, and are lost when that process terminates. Release builds do not save these tracking data as separate files or retain them as tracking history after restart. When a user starts a Camera only recording, camera images are saved as video frames under the storage, retention and deletion practices in Sections 3 and 5.2. Debug builds used for development and verification store diagnostic logs on the device, including detection counts, processing rates and some expression coefficients. These logs contain no camera images or full facial landmark arrays, and the app does not automatically transmit them. Stopping the camera stops new capture; in-flight processing may temporarily retain a frame until it finishes. Stopping does not mean that all memory is erased instantly.

5.2 Recordings containing facial renderings or actual faces

Avatar recordings, Camera only recordings, imported videos and edited copies have no automatic expiry. These in-app videos are retained until the user deletes them. In the iOS recording preview, choose Delete video using the trash icon to remove that video’s local original and edited copies from the app. The source video in Photos is unchanged. Close does not delete it. To remove all in-app recordings, including recordings whose previews were closed, use the system’s Delete App action to remove the app and its data. Offload App preserves data and does not remove all recordings.

Delete Photos and exported copies separately in Photos or Finder. Follow the respective app’s instructions to permanently remove copies in Recently Deleted or Trash. Manage and delete backup and shared copies separately through those services. Deleting in-app recordings or uninstalling the app does not delete these copies.

The developer cannot access face data or recordings on your device and cannot delete them remotely. For deletion assistance, use the contact in Section 1 without attaching footage.

6. Support requests

If you email us, we use your email address, your name if provided and the content of your request only as needed to respond and investigate. We retain this information for as long as necessary to handle the request, and delete it when no longer needed. You may request deletion using the contact in Section 1.

You do not need to attach face photos, camera footage or skeleton videos. We do not request face data as support material.

7. This website

This information site is hosted on GitHub Pages. The developer has not added advertising or analytics scripts. GitHub processes and logs information such as IP addresses for security when you visit. This website has no feature for uploading face images, facial landmarks or recordings.

See the GitHub General Privacy Statement for details.

8. Changes to this policy

If features or data handling change, this page will be updated with the applicable version and revised date.