Face Explorer Privacy Policy

Version: 2.5 Effective date: November 13, 2026 Applies to: Face Explorer (脸型探索) for iPhone, iPad and macOS Public URL: https://z415073783.github.io/app/FaceAnalyse/en/privacy.html Contact: z674636160@163.com


Summary

Face Explorer does not require an account. The photo, birth information, name, MBTI type or tarot question you choose to provide is sent over an encrypted connection to our own server. The server computes deterministic facts in memory (e.g., Four Pillars/Five Elements, zodiac, tarot cards, life path number, synastry relations) and a third-party AI service (DeepSeek) turns them into an entertainment reading. We do not store your photo, birth information, name, question or report text on our servers, and we do not use them for advertising or model training. We use only a randomly generated device identifier for request verification, daily limits and abuse prevention. For two-person scenes, you must obtain the other person's consent. Completed readings are saved only on your device (up to the latest 50, including facts, full report and a compressed thumbnail; no original image). You can delete/clear them at any time; we do not upload local history or thumbnails. Optionally, you may save a local-only “My Profile” (birth information, name and MBTI type) on your device to pre-fill readings and daily content; it is never uploaded and can be edited or deleted at any time in Settings (see Section 13).

On iOS, product usage statistics and crash diagnostics are provided by Firebase (Google) — Firebase Analytics and Firebase Crashlytics — and are initialized only after you consent to this policy; before consent, the app makes no network requests. Firebase is not used for advertising or ad personalization, we do not use IDFA, and we do not track you across apps (see Section 3 and the Appendix).


1. Information We Collect

1.1 Provided by you

1.2 Generated automatically

1.3 Purchases

1.4 What we do NOT collect

We do not collect account credentials, phone number, email (unless you email us), contacts, precise location, other photos in your library, IDFA/advertising identifiers, Face ID or other biometric templates, or information about people you share reports with. We do not integrate advertising SDKs, do not use IDFA or other advertising identifiers, and do not track you across apps. On iOS, product usage statistics and crash diagnostics are provided by Firebase (Google) and are initialized only after you consent to this policy (see Section 3).

Name, birth date/time/place/longitude, gender, MBTI type, tarot question and two-person information are processed only when you start the relevant reading, are not stored on our servers, not used for profiling, advertising or model training, and are not requested when you do not use those scenes. The only exception is the optional local profile you choose to save on your device (Section 13): it stays on that device and is never uploaded, and you can delete it at any time in Settings.

No network requests before consent: the app makes no network requests until you consent to this policy. On iOS, Firebase Analytics and Firebase Crashlytics are initialized only after consent and are not used for advertising or ad personalization. Firebase processes data under Google's policies (Google Privacy Policy, Firebase Privacy and Security); retention and cross-border processing of this data are governed by Google's policies. You can stop this collection by not consenting or by uninstalling the app (see Section 6).


2. How Photos Are Processed

  1. Only the photo you select: the app does not read your library in bulk. It uses the system photo picker or camera to minimize permission scope.
  2. Local compression and metadata removal: before upload, the photo is redrawn as JPEG (long edge about 1280 px, quality about 0.8), orientation is baked into the pixels, and EXIF, GPS location and camera-device metadata are removed.
  3. Encrypted transfer: the photo is base64-encoded in an HTTPS request to our own server (hk.zengliangmin.cn), protected by request signatures against tampering and replay.
  4. Forwarded to a third-party AI: our server forwards the photo to DeepSeek for visual analysis. See DeepSeek's privacy policy: Chinese / English. We forward data in a minimized form; the provider's retention, training and processing practices are governed by its own policy, and we make no commitments beyond it.
  5. In-memory processing, no persistence: our server validates, forwards and releases the photo in memory only. We do not write photos to server disks, do not use object storage, do not log them and do not back them up. Memory is released when analysis ends or the request is closed. Report text is streamed to your device and is not stored on our servers.
  6. Photos are not stored long-term on the device: the app keeps the photo in memory only for the current session; it is not written to your library or app cache. Local history stores result facts, the full report and one compressed thumbnail (240–320px long edge JPEG, EXIF removed, not the original); deleting history or clearing all removes thumbnails. Photos taken with the camera are processed the same way as library photos: used only for this analysis, the original is not stored on our servers; local history may keep a compressed thumbnail, which you can delete at any time.
  7. No training or advertising: we do not use your photos to train AI models, for advertising, for sale, or for sharing with research institutions.
  8. Report sharing: the share card/text is generated on-device only when you tap Share. It contains only the displayed report excerpt, scene, date, brand and disclaimer; it does not include birth details, names, questions, photos, two-person information or raw input. It is not uploaded or posted automatically; the destination third-party app handles it under its own privacy policy.

3. Third-Party Services

Third party Purpose Data involved Notes
DeepSeek AI entertainment reading inference (face/palm/Bazi/horoscope/birthday/tarot/name/MBTI/synastry) Compressed photo, birth information/name/MBTI/tarot question/two-person information (as applicable), report language Processing governed by DeepSeek's official privacy policy; servers may be located in mainland China or elsewhere, which may involve cross-border transfer
Firebase Analytics (Google) Product usage statistics and feature improvement (iOS only) Usage events, sessions, device model/OS version/app version, app instance ID, coarse region derived from IP Not used for advertising or ad personalization; no IDFA and no cross-app tracking. Processing governed by Google's privacy policy and Firebase Privacy and Security; initialized only after you consent to this policy
Firebase Crashlytics (Google) Crash and performance diagnostics (iOS only) Crash stack traces, device state, OS/app version, installation identifier Same processor and policies as above; initialized only after you consent to this policy
Apple (App Store / StoreKit) In-app purchases and receipts Transaction information Processed by Apple under its privacy policy; we do not access payment credentials
Our own server and CDN/reverse proxy API access, HTTPS, rate limiting Request metadata, network logs Used for service operation, security and abuse prevention

We do not sell your personal information or provide photos to third parties other than the services above, unless required by law or necessary to protect users or public safety. On iOS, Firebase data is not used for advertising, ad personalization or cross-app tracking.


4. Retention of Device ID and Usage Data


5. Data Security


6. Your Rights and Choices


7. Minors

The app is rated 12+ and is not designed for children under 12. If you are under 18, please use it with the consent and guidance of a parent or guardian. We do not actively collect age or date of birth. If a guardian believes a minor uploaded a photo without consent, contact us and we will help handle the relevant records.


8. AI-Generated Content and Disclaimer

Reports are generated by a third-party AI model based on the information you provide and traditional/entertainment rules, for entertainment and self-reflection only. They are not medical, investment, legal, psychological, marital/divorce or other professional decision advice. AI results may be inaccurate, incomplete or inconsistent with reality. Do not make medical, aesthetic, investment, legal or relationship decisions based on a report, and do not use it to evaluate other people or make decisions for them. Bazi/astrology/tarot/birthday/name/MBTI/synastry content is traditional or pop-culture entertainment, does not predict the future and is not guaranteed to be accurate; 12+. For two-person scenes you must obtain the other person's consent. If a photo/input cannot be analyzed (e.g., no clear face or palm, content filter), the app will ask you to change it and will not deduct a credit.


9. Changes to This Policy

We may update this policy due to feature changes, legal requirements or third-party service changes. The new version will be published on this page with a new effective date; material changes will be communicated in the app in a reasonable way. Continued use means you accept the updated policy.


10. Contact Us


11. Multi-Scene Entertainment Reading Data (New in v2.0)

  1. Scenes: face, palm, Bazi/Five Elements, horoscope, birthday, tarot, name, MBTI, synastry and Bazi compatibility, plus the existing classic face score.
  2. Server-side deterministic calculation: birth information is used to compute Four Pillars/Five Elements/true solar time in memory; tarot cards are drawn server-side; life path number/zodiac/synastry relations are computed by fixed rules. These facts and inputs exist only in memory for the request and are not written to disk or logs.
  3. Third-party AI: the server forwards the necessary facts and inputs (photo/birth information/name/question, as applicable) to DeepSeek to generate the entertainment reading. Processing by DeepSeek is governed by its official privacy policy; we require minimization and prohibit advertising use.
  4. Two-person information: you must confirm that you have obtained the other person's consent. We use it only for this reading, do not store it and do not use it for identification or profiling. If the other person contacts us, we will explain the processing; because we do not retain the data, it may not be possible to locate a specific request.
  5. What we do not do: no identity verification, age verification, medical/dermatological diagnosis, investment advice, marriage/divorce decisions, psychological diagnosis, advertising profiles, or cross-app tracking; we do not store birth information, name, MBTI, tarot questions, photos, two-person information or report history on our servers; completed readings may be kept only in local history on your device (Section 12), and an optional profile may be kept only on your device (Section 13).
  6. Age: 12+; not directed to children under 12. For third-party information you must ensure lawful permission.
  7. App Store privacy labels: "Other User Content" (birth information/name/MBTI/question) is collected for App Functionality, not linked to identity and not used for tracking; photos remain "Photos"; device ID is used for App Functionality/abuse prevention.

12. Local History, Thumbnails and Notifications (New in v2.1; updated in v2.4)

  1. Local history content: successful readings are saved on the device with scene, completion time, language, facts, full report and one compressed thumbnail (240–320px long edge JPEG, EXIF removed for image scenes; scene icon placeholder for text scenes). We do not store the original image, photo base64, raw input objects (birth date, name, tarot question, two-person raw input); derived display facts (e.g., solar/lunar date, Four Pillars, Five Elements) may be stored locally with the result; names/questions naturally appearing in the report are stored locally only as part of that report.
  2. Device-only, not uploaded: history and thumbnails live in the app's Application Support, are not sent to our server or third-party AI, are not synced across devices, and are excluded from iCloud backup.
  3. Limit and deletion: keep at most the latest 50 items; adding the 51st removes the oldest; individual delete and clear-all are supported with confirmation and are irreversible; deleting history cancels its pending local notification.
  4. Notification permission and purpose: before the first reading we ask for notification permission to remind you when a reading completes; optionally, you can enable daily fortune reminders in Settings. Declining does not block analysis or daily content, and only in-app banners are used where applicable. Both kinds of notifications contain only generic text — completion reminders show the scene name and a generic “tap to view” message; daily reminders show only a generic “today’s fortune is ready” message — no birth information, name, question, MBTI type, report text or thumbnail; you can turn completion alerts and daily reminders off at any time in the app or in system settings.
  5. Billing separation: saving/viewing/deleting history does not refund, recharge or consume credits; "read again" starts a new reading and is charged as a new reading.
  6. Birth information/name/MBTI/tarot question/two-person information: still processed in server memory under this policy, not written to disk/logs and no server-side history; local history does not additionally store raw inputs.
  7. Age: local history/notifications are also 12+ entertainment; not used for advertising, profiling or cross-app tracking.

13. Local Profile, Daily Content and Reminders, Report Cache (New in v2.4)

  1. Local profile (optional, device only): “My Profile” lets you optionally save birth information, name and MBTI type on your device so that readings and daily content can pre-fill them. It is stored in the app’s Application Support, excluded from iCloud backup, never uploaded to our servers or the third-party AI, and not included in local history as an input record. You can edit or delete it at any time in Settings; deleting it does not affect saved history or credits. If you never create a profile, no profile data is stored.
  2. Daily content: when you view the daily fortune card, the app sends POST /api/v1/daily to our server with either a zodiac sign or your birth month and day (derived on the device) plus the interface language. The server computes deterministic facts (e.g., sign, lucky color/number, daily tarot card) in memory and returns them. No third-party AI is called, no analysis credit or quota is consumed, and the endpoint does not create server-side history linked to you.
  3. Daily reminders (optional, local): if you turn on daily reminders, the app schedules local notifications at 08:30 and 21:00 in your device’s time zone. Their content is generic (“today’s fortune is ready”); no birth information, name, MBTI type, question, report text or thumbnail is included. Reminders are scheduled and shown by the system on your device; you can turn them off at any time in Settings or system settings. Declining or disabling them does not affect daily content or readings.
  4. MBTI quiz: the 12-question personality quiz is scored locally on your device. Your answers are not uploaded; only if you choose to save the resulting 4-letter type to your local profile, or start an MBTI reading, is the type stored on the device / sent as the reading input under Section 1.
  5. Server-side generic report cache: for horoscope reports and daily tarot (single card, no question), the server may cache the generated report. The cache key contains only non-identifying parameters (sign and period/date, language, prompt version); it contains no device identifier, photo, birth date, name, MBTI type, question or other personal data, and cannot be used to build a user profile. Cached entries expire at the end of the period + 1 hour (daily tarot: next midnight + 1 hour) and the cache holds at most 500 entries, evicting the oldest. Personalized tarot readings with a question are not cached.
  6. Rating prompt: after two consecutive days of use, the app may ask the system to show the standard App Store rating prompt at most once. The decision is made locally on your device; we do not collect or receive any data about whether the prompt was shown or what rating you gave (Apple processes it under its own policy), and the prompt can be dismissed without affecting the app.

Appendix: App Store Privacy Label Summary (for review and development alignment)

Data type Collected Purpose Linked to identity Used for tracking
Photos or videos (user content) Yes (uploaded only when you start an analysis) App functionality No No
Device ID (random UUID) and Firebase app instance ID (iOS only) Yes App functionality, abuse prevention; app instance ID used for usage analytics on iOS No No
Usage data (request/usage metadata) Yes (server-side counters) App functionality, analytics (cost and reliability) No No
Usage data (Firebase Analytics, iOS only: usage events, sessions, device and OS/app version, coarse region) Yes (iOS only, after you consent) App functionality, analytics (feature improvement) No No
Purchases Handled by Apple; credits stored locally App functionality No No
Diagnostics (Firebase Crashlytics, iOS only: crash stack traces, device state, OS/app version, installation identifier) Yes (iOS only, after you consent) App functionality, analytics (crash and performance diagnostics) No No
Other user content (birth information/name/MBTI/question/two-person information) Yes (only when you start the relevant scene) App Functionality No No
Local profile (birth information/name/MBTI, on-device only) Yes (only if you save it) App Functionality No No
Local history/thumbnails (device only) Yes (stored locally only, not uploaded) App Functionality No No
Notification permission/local notifications Yes (completion reminders + opt-in daily fortune reminders) App Functionality No No

Note: the Firebase rows apply only to the iOS version; the macOS version does not integrate Firebase and does not collect this usage or diagnostics data.