Nest Identify/Auto-Match/Modify under /people with redirects; reject
blur/extreme pose at Process; Name-cluster on Identify; axe smoke tests;
helpers to rotate DEV admin password and seed match_decisions.
Add light/dark toggle, navy gold-accented sidebar, skip link, and
clearer titles so admin reads as the same JRCC brand; document UX
review and next priorities in ROADMAP.
Admin brand PNGs were gitignored so /brand/* 404'd to SPA HTML;
track the same assets as the viewer and map gray/blue page classes
to semantic theme tokens so admin matches viewer styling.
POST /people/{keep_id}/merge reassigns faces, videos, and match
decisions, rebuilds encodings, then deletes sources so strict
Auto-Match thresholds stay recoverable.
Stop shipping punimtag_password / admin defaults in install helpers and
docs; generate secrets at install time; require ADMIN_PASSWORD from env
(with test-only bootstrap in conftest).
Port viewer design tokens to admin shell/login with JRCC branding; add max recognition distance, next-best person margin, stricter detection floor, and tighter auto-accept distance.
Raise reference quality floor to 0.5, align browse/run/auto-accept defaults with UI copy, and persist match_decisions for later confidence calibration.
Capture optional contact fields on identify/approve, add Select All and
Approve Next 10 on User Identified Faces, and email admins when the
pending queue grows (ADMIN_NOTIFY_EMAIL via existing noreply SMTP).
- #45: re-read local folder when recursive checkbox toggles
- #47: clear Identify crop spinner when image is already cached
- #46: search selected people by person_ids; AND for "First Last"
- #21: enqueue network import without blocking API walk
- #26: serve resized grid thumbnails for photos
- #30/#31/#33: re-enable small-face filters in auto-match
Add on-demand H.264/AAC web playback (RQ, ffmpeg) with API routes and
Next.js proxies; extend admin UI with WebPlaybackVideo and shared hooks.
Store transcode cache beside pending-photos (WEB_VIDEO_CACHE_DIR / UPLOAD_DIR)
and ignore data/web_videos. Centralize FastAPI URL helpers, optional Vite
and Next base paths for subfolder deploy, and fix modal reopen by using
router.replace when closing the home photo viewer. Include migration,
install scripts, deployment doc updates, and CI admin build env tweak.
Made-with: Cursor
- Remove unused isPlaying state that was causing TypeScript error TS6133
- Component only needs showPlayButton to control play button overlay visibility
- Fixes build error: 'isPlaying' is declared but its value is never read
- Show confirmation dialog before starting auto-match operation
- Inform users that it's a bulk operation processing entire photo library
- Warn that some matches may not be 100% accurate
- Recommend reviewing results after completion
- Create VideoPlayer component with centered Play button overlay
- Add /video/:id route to display videos with Play button
- Update Search page to open videos in VideoPlayer page instead of direct URL
- Play button shows when video is paused/stopped and hides when playing
- Button includes hover effects and doesn't interfere with native controls
- Introduced a new input field for entering new tag names alongside selecting existing tags.
- Updated the handleAddTag function to support adding multiple tags at once.
- Improved user experience with alerts for empty tag submissions and success feedback.
- Enhanced UI with clear labeling and instructions for tag management.
- Replace localStorage-based dev mode with VITE_DEVELOPER_MODE env var
- Remove setDeveloperMode function and localStorage logic
- Update Settings page to show status instead of toggle
- Add VITE_DEVELOPER_MODE to vite-env.d.ts type definitions
- Update .env_example with developer mode documentation
Developer mode is now controlled by setting VITE_DEVELOPER_MODE=true
in admin-frontend/.env file (requires rebuild to take effect)
- Add debug mode support for encoding statistics in API responses
- Debug info includes encoding length, min/max/mean/std, and first 10 values
- Frontend logs encoding stats to browser console when debug enabled
- Identify page enables debug mode by default
- Implement distance-based confidence thresholds for stricter matching
- Borderline distances require higher confidence (70-95% vs 50%)
- Applied when use_distance_based_thresholds=True (auto-match)
- Reduces false positives for borderline matches
- Dual tolerance system for auto-match
- Default tolerance 0.6 for regular browsing (more lenient)
- Run auto-match button uses 0.5 tolerance with distance-based thresholds (stricter)
- Auto-accept threshold updated to 85% (from 70%)
- Enhance pose detection with single-eye detection
- Profile threshold reduced from 30° to 15° (stricter)
- Detect single-eye visibility for extreme profile views
- Infer profile direction from landmark visibility
- Improved face width threshold (20px vs 10px)
- Clean up debug code
- Remove test photo UUID checks from production code
- Remove debug print statements
- Replace print statements with proper logging
backend/config.py - Added MIN_AUTO_MATCH_FACE_SIZE_RATIO = 0.005
backend/services/face_service.py - Multiple changes:
Added load_face_encoding() function (supports float32 and float64)
Added _calculate_face_size_ratio() function
Updated find_similar_faces() to filter small faces
Updated find_auto_match_matches() to exclude small reference faces
Fixed reference face quality calculation (use actual quality, not hardcoded 0.5)
Fixed duplicate detection (exclude faces from same photo)
Updated confidence threshold from 40% to 50%
Updated confidence calibration (moderate version)
backend/api/faces.py - Updated default tolerance to 0.5 for auto-match endpoints
backend/schemas/faces.py - Updated default tolerance to 0.5
admin-frontend/src/pages/AutoMatch.tsx - Updated default tolerance to 0.5
admin-frontend/src/api/faces.ts - Added tolerance parameter support
- Added functionality to create new tags and update existing tags in bulk.
- Implemented local state management for tags to improve user experience.
- Updated UI to allow users to enter new tag names alongside selecting existing ones.
- Ensured tags are reloaded in the parent component after creation for synchronization.
- Backend: Updated list_people_with_faces to search by first_name, middle_name, last_name, and maiden_name
- Frontend: Updated Modify page search UI and API client to support extended search
- Frontend: Updated Help page documentation for new search capabilities
- Infrastructure: Added start-api.sh wrapper script to prevent port 8000 binding conflicts
- Infrastructure: Updated PM2 config with improved kill_timeout and restart_delay settings
- Added new logging scripts for quick access to service logs and troubleshooting.
- Updated job streaming API to support authentication via query parameters for EventSource.
- Improved photo upload process to capture and validate EXIF dates and original modification times.
- Enhanced error handling for file uploads and EXIF extraction failures.
- Introduced new configuration options in ecosystem.config.js to prevent infinite crash loops.
- Local mode: Use File System Access API (Chrome/Edge/Safari) or webkitdirectory (Firefox) to read folders from browser
- Local mode: Browser reads files and uploads them via HTTP (no server-side filesystem access needed)
- Network mode: Type network paths or use Browse Network button for server-side scanning
- Add 'Start Scanning' button for local mode (separate from folder selection)
- Update API client to handle FormData uploads correctly (remove Content-Type header)
- Update Help page documentation with new scan mode options
- Add progress tracking for local file uploads"
- Add `browseDirectory` API endpoint to list directory contents.
- Create `FolderBrowser` component for user interface to navigate directories.
- Update `Scan` page to integrate folder browsing feature.
- Define `DirectoryItem` and `BrowseDirectoryResponse` schemas for API responses.
- Added media_type to PhotoSearchResult interface to distinguish between images and videos.
- Updated PhotoViewer component to support video playback, including URL handling and preloading logic.
- Modified openPhoto function in Search page to open videos correctly.
- Enhanced backend API to serve video files with range request support for better streaming experience.
These changes improve the user experience by allowing seamless viewing of both images and videos in the application.
- Update API client to use relative paths when VITE_API_URL is empty
- Fix EventSource URLs to use window.location.origin for proxy compatibility
- Update image/video URL construction to use relative paths
- Add debug logging for API response troubleshooting
- All API calls now work correctly when served through HTTPS proxy
This fixes mixed content errors and allows the admin frontend to work
when accessed via HTTPS domain with reverse proxy (Caddy/nginx).
- Added a new API method to fetch photo images as blobs, enabling direct image retrieval.
- Updated image source paths in multiple components to use the base URL from the API client for consistency.
- Implemented cleanup for blob URLs in the ReportedPhotos component to prevent memory leaks.
- Improved user experience by displaying loading states for images in the ReportedPhotos component.
These changes improve the efficiency and reliability of photo handling in the admin interface.
This commit introduces a new `DEPLOYMENT_CHECKLIST.md` file that outlines the necessary steps for configuring server-specific settings after pulling from Git. It includes instructions for environment files, PM2 configuration, firewall rules, database setup, and building frontends. Additionally, it adds an example `ecosystem.config.js.example` file for PM2 configuration, providing a template for users to customize for their deployment environment. The `.gitignore` file is updated to include the new PM2 ecosystem config file.
This commit updates the Python linting rules by adding additional flake8 error codes to ignore in both the CI workflow and the Python linting command in package.json. It also modifies the ESLint configuration for the admin frontend to streamline the linting process by removing the max-warnings restriction.
This commit enhances the linting configurations by adding additional flake8 error codes to ignore in both the CI workflow and the Python linting command in package.json. It also modifies the ESLint configuration for the admin frontend to remove the report for unused disable directives, streamlining the linting process and reducing false positives.
This commit updates the CI workflow to include steps for auditing dependencies in both the admin and viewer frontends, ensuring that vulnerabilities are identified and addressed. Additionally, it updates the `package-lock.json` and `package.json` files to reflect the latest versions of `vite` and other dependencies, improving overall project stability and security. These changes contribute to a more robust development environment and maintain code quality.
ilia
Tanya