LocalCaptions: Privacy Policy
Privacy Policy
Application: LocalCaptions
Developer / Organization: REVAMINE TECHNOLOGIES
Contact Email: socialmail132@gmail.com
Last Updated: October 2026
Core Privacy Commitment (Executive Summary)
Your privacy is absolute. All video transcription and subtitle generation happens 100% locally on your device using offline neural speech models. We do not operate external servers to collect, upload, or analyze your personal audio, video, or subtitle text.
1. Information We Do NOT Collect
We strongly believe in minimal data footprint. We do NOT collect:
- No Account Information: No registration, login, email address, phone number, or passwords are required to use the app.
- No Media Uploads: Your chosen video and audio files are processed entirely inside your phone's memory and CPU. They are never transmitted to any remote cloud server.
- No Subtitle Logging: Transcribed text, SRT files, and project names are stored strictly on your local device storage.
- No Identity Tracking: We do not collect advertising IDs, biometric data, contacts, or GPS location.
2. Device Permissions Used & Why
The application strictly requests only the minimal permissions required for essential functionality:
| Permission | Purpose | Data Handling |
|---|---|---|
| RECORD_AUDIO | Used strictly when you tap the live microphone to dictate or transcribe live speech in real time. | Audio is processed on-the-fly by the on-device AI engine. Audio streams are never stored remotely or shared. |
| INTERNET | Used only to download offline speech models (weights) selected by you from our repository, and optional external AI APIs if enabled by you. | No user media or analytics are uploaded over this connection. |
| Photo & Video Picker | Enables you to select videos or audio tracks from your phone to generate subtitles. | Uses Android's secure Photo Picker. The app does not request or require broad storage access. |
3. AI Speech Models & Third-Party Services
To keep the initial download size small, our app lets you download speech-to-text models on demand:
- Model Hosting: Speech recognition models (INT8 ONNX format) are hosted on Hugging Face (Repository:
NiranjanYadavIN/indicconformer-sherpa-onnx, originally packaged byparismitaglobalsolutions). When you tap "Download", the model weights and vocabulary files are downloaded over secure HTTPS directly to your private app storage. - Open Source Technologies: We utilize open-source speech engines including AI4Bharat IndicConformer (MIT), OpenAI Whisper (MIT), NVIDIA NeMo (CC-BY-4.0), and Next-gen Kaldi / sherpa-onnx (Apache 2.0).
- Optional Cloud AI (Gemini): If you choose to enter your own Gemini API key in Settings for AI enhancement, requests are sent directly to Google APIs using your own key. If no key is entered, no cloud AI calls are ever made.
4. Data Storage & User Control
Because all data is stored locally in your app's internal database (SQLite Room):
- Deletion: You can delete any project, subtitle segment, or downloaded speech model at any time from within the app.
- Cache Clearing: You can clear downloaded files and temporary caches via the Settings menu or Android App Info.
- Uninstallation: Uninstalling the application immediately deletes all local databases, projects, and downloaded models from your device.
5. Children's Privacy
Our application does not knowingly collect, store, or solicit personal information from children under the age of 13 (or under 16 in the European Union). Since the app does not collect personal data from any user, it is compliant with COPPA and global child privacy regulations.
6. Changes to This Privacy Policy
We may update this Privacy Policy from time to time to reflect app improvements or regulatory updates. Any changes will be posted on this page with an updated revision date.
7. Contact Information
If you have any questions, concerns, or requests regarding this Privacy Policy, please contact us:
REVAMINE TECHNOLOGIES
Email: socialmail132@gmail.com