Timeline, Memories, and Duplicates now share a single fixed cell size
(THUMBNAIL_SIZE=160) with no 1fr stretching — cells stay exactly 160px
regardless of sidebar state, at the cost of a small right-edge strip
when the container width isn't a multiple of (160+gap). Width is
measured on the scroll container itself with padding subtracted so
sidebar expand/collapse reliably reflows the grid.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Timeline's items array used to rebuild on every sub-pixel cellSize tick
during the sidebar CSS transition, causing visible jank with thousands
of photos. Row heights now resolve off a ref at virtualizer-measure
time, so items only rebuild when the column count actually changes.
PreviewFilmstrip is horizontally virtualised (~15 cells in the DOM
instead of N), cutting preview open latency on large libraries. Also
honor the user's explicit right-sidebar collapse (don't auto-reopen on
photo selection) and shrink the sonner toasts to a tighter form factor.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Memories view now supports the same keyboard shortcuts, heap membership,
and optimistic cache updates as the Timeline. Arrow/Ctrl+A/Escape nav is
extracted into a shared useGridKeyNav hook so both views stay in lockstep.
Duplicates view is virtualised with @tanstack/react-virtual and has
stabilised PhotoThumbnail props so React.memo actually elides work when
scrolling or toggling selection.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
MemoriesView now renders PhotoThumbnail cells wired up to the shared
photoStore so selection, heap membership, preview (double-click /
Enter), badges, drag-to-heap, and search-match highlighting all work
the same way they do in Timeline. Kept the per-year section grouping,
swapped the bespoke img tiles for the shared component.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace the free-text username input with a Select populated from a new
/sharing/users endpoint. Users already on the target's share list are
filtered out, and the trigger surfaces loading / empty states. Matches
the existing permission model since sharing only ever required knowing
a username.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Perf / a11y (high-impact review items)
- Timeline arrow-key handler binds once per (viewMode, currentSection)
and reads fresh state via navStateRef instead of an 8-element dep
array of new-each-render values.
- usePhotosQuery collapses 14 individual Zustand selectors into one
useShallow selector returning the params object.
- PhotoThumbnail no longer subscribes to the search query directly;
Timeline subscribes once and passes it down as a prop.
- PhotoThumbnail gains role="button", tabIndex, aria-label, aria-pressed,
Enter/Space key handlers and a focus-visible ring. Timeline marked
role="grid"; RightSidebar marked role="region".
Style consistency
- Swap clsx for cn (tailwind-merge aware) across 17 files so
conflicting utility classes collapse correctly.
- New Badge primitive (ui/badge.tsx) with default/neutral/overlay/
outline variants; adopted in ColorsView, RatedView, TagsView for
the repeated count overlay pill.
- Fix palette drift: text-amber-400 -> text-star, text-green-*
-> text-pick, text-red-* -> text-reject (5 files).
- Button gains an xs size (h-6 px-1.5 text-[11px]) for the repeated
compact-button pattern.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Frontend cleanup pass driven by the post-shadcn review.
Performance
- Memoize PhotoThumbnail and route cell click/double-click through
stable handlers so heap-membership invalidation no longer re-renders
every visible thumbnail.
- Cap usePhotosQuery's eager background page-walk at 20 pages with a
50ms inter-page yield — was unbounded (up to 100k photos cold).
- Drop the per-thumbnail loading spinner in favour of the existing
pulse skeleton; only retry state still surfaces a spinner.
UX
- Coalesce rapid X/U presses into a single undo entry + one toast
(1.2s window) so accidental bursts are easy to back out.
- Optimistic rating/color updates with per-id snapshot rollback on
error, matching the existing discard pattern.
- Section-aware empty timeline state with a Clear-all-filters CTA.
- Carry the search-match chip from the grid into the preview header.
- Add a basket-icon badge for active heap membership so the green
tint isn't the only signal (colorblind-safe).
- Standardise error toasts via formatApiError(): FastAPI detail,
validation arrays, axios message, with a 'Network Error' filter.
Architecture
- Extract useBulkPhotoMutations and stop duplicating
bulkRating/bulkColor across RightSidebar and useKeyboardShortcuts.
- Split RightSidebar (714 -> 448 LOC) and PhotoInfoPanel (952 -> 716)
into co-located sub-components: BulkTakenAtEditor, BulkTagsEditor,
TagsEditor, TakenAtEditor.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Adopts shadcn/ui components (Dialog, Button, Input, Select, Popover,
Command, Checkbox, Switch, Toggle, Calendar, etc.) across the app,
replacing hand-rolled modals, dropdowns, and form controls. Adds a
reusable cmdk-backed MultiSelect for the Type, Tags, and Flag filters
so all multi-value filter popovers share one component and layout.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Selection now reads as a blue ring + tint with a springy scale-down,
hover stays a subtle gray ring so keyboard-driven and mouse-driven
states are tellable apart.
- Heap membership is signalled with a green tint only (no badge, no
ring, no scale).
- Discard/restore is optimistic and non-yanking: photos stay in the
grid greyed out until the next reload, X toggles based on the
current state, and the same treatment applies in preview.
- Filmstrip mirrors the grid styling (selection blue, heap green,
discarded grey).
- Preview close restores the LAST viewed photo as the focused/selected
one in the grid.
- Right sidebar collapses on view change and re-opens when a photo is
in focus; Esc clears active selection so the panel collapses too.
- Keyboard hints panel is context-aware (grid / preview / discarded
section), collapsible with H, persisted, and rendered inside the
preview column above the filmstrip.
- "Pick (P)" renamed to "Select (S)" everywhere.
- Needs review moved into the Flag pill dropdown.
- Fixed vertical videos overflowing the preview column (min-h-0).
- Replaced the bottom-right ScanProgress popover with an inline
spinner next to the FOLDERS sidebar header (and on the specific
folder row being scanned). ScanProgress is now a headless
invalidator; useScanActivity exposes the live status.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Folder tree now indents 20px per level (chevron width + gap) so a child's
chevron column lines up under its parent's label, and depth-1 rows nest
under the section eyebrow instead of starting flush with it. Spacer for
leaf rows matches the chevron button footprint so rows align regardless
of expandability.
Sidebar open/close buttons (previously split between TopBar and each
panel header) collapse into two toggles at the ends of the FilterBar.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Drops face recognition, OCR, object detection, and semantic embeddings.
The sole remaining vision task is a CLIP-based binary classifier
(photography vs other); photos in "other" get needs_review=true so
screenshots, documents, memes and scans can be triaged from a new
filter pill in the UI.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Adds Redis-backed feature flags for vision stages with admin UI toggles
and manual backfill trigger, photo upload and download routers with
frontend upload modal, and rawpy-based RAW decoding with JPEG fallback
for misnamed DNGs. Fixes pgvector serialization, is_trashed filter, and
naive-datetime bind in incremental duplicate regrouping; bumps Celery
time limits on regroup tasks beyond the 5-minute default.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace 3K+ React CircleMarker components + MarkerClusterGroup with
native Leaflet L.circleMarker on a shared L.canvas() renderer added
in a single useEffect. Zero React components per marker — canvas
draws all points in one paint (<50ms vs multi-second freeze).
Also drops react-leaflet-cluster from the bundle (-46KB gzipped).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace DivIcon thumbnail markers with lightweight CircleMarkers.
Each DivIcon created a DOM element with an <img> tag, so 3K+
geotagged photos meant 3K DOM nodes and 3K thumbnail requests
hitting simultaneously — freezing the browser during clustering.
CircleMarkers are SVG-rendered on Leaflet's canvas layer with no
DOM nodes per marker. Photos still open in preview on click.
Also: bump cluster radius 50→80, enable removeOutsideVisibleBounds,
disable clustering at max zoom, increase staleTime to 5 min.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
HEIC thumbnails:
- Switch fallback from ffmpeg to vips for HEIC files that pillow-heif
rejects. ffmpeg decoded gain map tiles instead of the primary image,
producing inverted/negative thumbnails. vips uses libheif's item
references correctly and extracts the full-resolution primary image.
Database pool exhaustion:
- Add idle_in_transaction_session_timeout=60s so Postgres auto-kills
leaked connections from disconnected thumbnail requests.
- Add pool_timeout=10 so new requests fail fast instead of hanging.
- Bump pool from 5+5 to 10+10 for thumbnail concurrency headroom.
- get_db rolls back on exception before closing.
Auth session persistence:
- Narrow 401 interceptor exclusion to only /auth/refresh and /auth/login
(was excluding all /auth/* including /auth/me, preventing token refresh
on boot).
- fetchMe only clears tokens on 401/403, not network errors.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
iPhone photos with depth maps or gain maps have too many auxiliary
image references for libheif 1.17, causing pillow-heif to throw
"Too many auxiliary image references". process_heic_image() now
falls back to ffmpeg when pillow-heif fails — ffmpeg's own HEIC
decoder handles these files without issue. Fixes 27/35 HEIC photos
that were stuck in failed state.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Sharing:
- New HeapShare and FolderShare models with read/write permissions
- Sharing API router (CRUD for heap and folder shares)
- Heap endpoints accept shared access (photo_ids, add/remove with write)
- Photo list drops user_id filter in shared context, adds owner_username
- Media serving (thumb/original/proxy) falls back to share check on 404
- ShareDialog component for managing shares from kebab menus
- HeapsPanel shows "Shared with me" section for shared heaps
- LeftSidebar shows "Shared with me" section for shared folders
- Owner badge on PhotoThumbnail for photos from other users
Auth:
- Access token default bumped to 1 year, refresh to 10 years
- Refresh token persisted in localStorage (survives page reload)
- Timer-based refresh replaced with 401 axios interceptor
Vision pipeline fixes:
- Bootstrap sets Redis ready key even on partial export failure
- Export functions run conditionally (only for actually missing models)
- _load_thumb handles multi-user path (/data/thumbs/{user_id}/{photo_id}/)
- can_access_photo_via_share uses single subquery instead of N+1 loop
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Addresses 16 robustness, transparency, and performance issues across
the Celery media processing pipeline:
Critical:
- Singleton DB engine in vision tasks (was leaking one per task call)
- acks_late + task_reject_on_worker_lost so crashed workers don't lose tasks
- Global soft/hard time limits (5/10 min) to prevent hung worker slots
- Thumbnail copy-before-resize (in-place mutation degraded larger sizes)
- backfill_vision now checks each task type independently (OCR, faces, etc.)
- Parameterized LIMIT in backfill_vision (was f-string SQL injection)
High:
- try/except + retry(max=3) on all vision inference tasks
- extract_metadata writes processing_error on exiftool failure
- PIL Image handles closed in _load_thumb/_load_original
- Scan progress Redis keys auto-expire after 1 hour
- Watcher lock renewal is wall-clock based (30s) not event-count based
- worker_process_init signal warms up vision models on startup
Medium:
- Explicit task_routes for every task name (wildcards never matched)
- app.services.metadata added to Celery include list
- POST /maintenance/recover-stuck endpoint for photos stuck in processing
- Docker healthchecks for worker-light, worker-vision, and Redis
- Task ID in vision log lines for distributed tracing
- Bare except:pass narrowed to specific exceptions
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add active source root directories to the library stats endpoint and
display them in the settings page. Hardcode container PHOTO_DIRS to
/photos since the volume mount handles host path mapping. Add .env to
.gitignore to prevent committing secrets.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Every user — including the initial admin — now gets their own
subdirectory under PHOTO_DIRS (e.g. /photos/admin, /photos/bob).
No one's source root points to the mount root itself, eliminating
cross-user photo overlap entirely.
- Setup endpoint: admin gets /photos/{username} like everyone else
- Migration: default admin media_path set to /photos/admin
- Remove scan directory pruning (no longer needed)
- Fix thumbnail retry URL: use & separator when token query param
already present (was producing ?token=...?retry=N)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When the admin's source root is the mount root (/photos) and other
users have subdirectories (/photos/bob), the admin's scan now prunes
those directories from os.walk so photos aren't double-indexed under
the wrong user. The scanner queries all active source roots owned by
other users and excludes their paths during directory traversal.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- detect_objects, classify_content, recluster_faces now look up the
photo's user_id and set it on created Tag rows — fixes tags being
invisible to the owning user due to NULL user_id
- Initial admin setup creates source root at the mount root (/photos)
instead of a subdirectory, since the admin owns the entire library
- Revert to OpenCLIP ViT-B/32 (512-d) as default embedder — SigLIP
requires transformers version alignment not yet available in the
Docker image. SigLIP2 code remains for future enablement.
- Add transformers to requirements for future SigLIP support
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Move watch_folders to dedicated 'watcher' queue with its own
single-concurrency container so it never blocks scan/thumbnail slots
- Add get_current_user_media dependency that accepts ?token= query
param for <img src> / <video src> media endpoints (thumb, original,
proxy) — fixes 401 on thumbnails
- Append JWT token to all media URLs in the frontend
- Add missing 'memories' case in sidebar navigation switch
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Re-enable the watchfiles-based folder watcher with a Redis lock to
prevent multiple instances from stacking up across restarts. The
watcher is now automatically dispatched on startup when scanner.watch
is true (default), and only one instance runs at a time.
- Redis lock (SETNX + TTL renewal) ensures single-instance execution
- Graceful exit if another watcher holds the lock
- New POST /maintenance/start-watcher endpoint for manual control
- Fix: use settings.scanner/vision properties instead of mulita_config
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add a Memories view that surfaces photos taken on the current date in
previous years (like Google Photos / Immich). Only uses EXIF-sourced
dates to avoid false matches from filesystem timestamps.
- Backend: GET /api/v1/photos/memories returns groups by year, up to
12 photos each, filtered to non-discarded/non-hidden EXIF dates
- Frontend: MemoriesView with year-grouped thumbnail grid
- Sidebar: new "Memories" nav item with clock icon
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Centralize execution provider selection in providers.py with
auto-detection and graceful fallback. All ONNX sessions (embedder,
detector, face processor, recognizer) now use the configured providers.
- New VISION_EXECUTION_PROVIDERS env var: "auto" for GPU auto-detect,
or explicit "CUDAExecutionProvider,CPUExecutionProvider"
- Provider priority: CUDA > ROCm > OpenVINO > CPU (when set to "auto")
- docker-compose.yml includes commented-out NVIDIA GPU deploy section
- Supports onnxruntime-gpu as a drop-in replacement for onnxruntime
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace OpenCLIP ViT-B/32 (512-d, ~78% recall) with SigLIP2 ViT-B/16
(768-d, ~84% recall) as the default embedding model for significantly
better image-text retrieval quality.
- New SigLIP2Embedder class with 384px input and SigLIP normalization
- ONNX export pipeline for SigLIP2 visual + textual encoders
- Migration 0010: resize embeddings.vector from 512 to 768 dimensions
- Config-driven model selection: "siglip2_vitb16" (default) or
"openclip_vitb32" (legacy) — both models can coexist
- Content classifier follows the configured embedder family
- Existing embeddings cleared on migration; vision backfill regenerates
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace O(N²) pHash-only duplicate detection with a hybrid approach:
- pHash Hamming distance for exact/near-exact copies
- CLIP embedding cosine similarity via pgvector HNSW for visually
similar photos (crops, format changes, screenshots)
Post-scan now uses incremental mode: only newly added photos are
compared against the full library — O(new × log N) via HNSW index
instead of O(N²). Full regroup remains available from Settings.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace page-number walking with cursor chaining in usePhotosQuery.
Each response includes a next_cursor that seeks directly to the next
slice via an indexed range scan — O(1) regardless of depth instead of
OFFSET-based skipping that degrades on large libraries.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Introduce username/password authentication with admin and user roles.
Each user gets their own media directory under /photos/{username}/ with
isolated photos, folders, heaps, and tags. Admins manage users and
observe the full library from a dedicated Settings page.
Backend:
- User model with bcrypt passwords and JWT access/refresh tokens
- Auth router (login, refresh, setup, change-password, status)
- Admin router (user CRUD with last-admin protection)
- user_id FK added to photos, folders, source_roots, heaps, tags
- All data routers scoped by authenticated user
- Scanner inherits user_id from source root owner
- Thumbnails stored under user-prefixed paths for isolation
- Library endpoints accept ?scope=global for admin cross-user view
- Alembic migration 0009 with data migration for existing installs
- Defensive bootstrap.py handles fresh vs existing DB startup
Frontend:
- AuthContext with token lifecycle, auto-refresh, login/logout
- Login page, first-run setup page, auth gate in App.tsx
- Bearer token interceptor on all API requests
- User identity + logout in left sidebar
- Admin-only Settings page with Library Management and Users tabs
- UserManagement panel (add, edit role, reset password, deactivate)
- Settings shows global stats across all users for admin
- Filter bar, right sidebar, keyboard hints hidden on settings page
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Pool was 20+10 overflow per engine; uvicorn --reload leaked pools until
Postgres hit max_connections=100. Reduced to 5+5 with pool_pre_ping.
Grouped views (tags, people, colors, rated) now sort cards largest-first.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Scan now automatically queues backfill_vision (+90s) and recluster_faces
(+300s) after dispatching folder scans. Face extraction also schedules a
debounced recluster via Redis so incremental file-watcher imports get
clustered without manual intervention.
The ScanProgress widget now tracks worker queue activity beyond the scan
phase, showing a "Processing Photos" indicator with vision queue counts
while background tasks (embeddings, faces, tags, OCR) are running.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Renders a thin top-of-cell banner on each photo while a text search
is active, labelling which metadata field matched (filename, title,
note, tag, or EXIF key name) and showing a short excerpt with the
exact matched substring highlighted in amber. Extends the list
endpoint's search to also match photos whose tag names contain the
query so tags show up alongside filename/title/notes/EXIF hits.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Tracks the scroll position and shows a small date chip pinned to the
right edge of the timeline, fading in while the user scrolls and out
700ms after they stop. Only active in date-sorted views — other sort
modes hide it since the label would be meaningless. A cached
row-offset/date index keeps the lookup to a single binary search per
scroll frame.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Rolls the standalone "Date issues" toggle back into FlagFilter as a
third value ('any' | 'discarded' | 'date_warning') so the date-warning
control lives in the same popover as Discarded, where operators expect
all flag-style filters. Drops the redundant dateWarning boolean and
its URL param.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Lets operators fix corrupted capture dates at scale. Adds an editable
Date Taken field with a folder/filename-derived suggestion hint, a bulk
Date Taken section in the multi-select sidebar that either applies one
date to the whole selection or infers a per-photo date from each path,
a warning badge on thumbnails whose stored date disagrees with the
path, and a "Date issues" filter pill so suspicious photos can be
surfaced and fixed as a group. Edits are written back to EXIF on disk
so rescans don't clobber the fix.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Adds a per-folder "hide from views" toggle so noisy subtrees
(screenshots, WhatsApp dumps, work archives) can be excluded from
cross-cutting views without losing indexing. Photos under a hidden
folder are still scanned, thumbnailed, embedded, OCR'd, face-
extracted — they just stop appearing in All Photos, Rated, Colors,
Map, Tags, People, Search, Duplicates, and the sidebar counts.
Navigating directly into the folder still shows every photo.
Schema (migration 0007_folder_hidden):
- folders.is_hidden user-set toggle, default false
- photos.is_hidden denormalized effective flag (true iff any
ancestor folder is hidden), indexed so cross-
cutting queries stay on the existing planner
paths
The denorm is maintained by two paths:
- The scanner walks the ancestry chain on insert, with a per-scan
memoized cache so each folder is resolved once per scan.
- POST /api/v1/folders/{id}/hide flips folders.is_hidden and runs a
WITH RECURSIVE CTE to recompute every folder's effective state in
one query, then bulk-updates photos WHERE IS DISTINCT FROM. Runs
in ~10 ms on a 13k-photo library.
Filters added (cross-cutting queries):
- /library/stats — every sidebar badge via a shared `visible` filter
- /photos (list) — only when neither folder_id nor heap_id is set;
folder browse and heap browse always show everything
- /photos/map
- /library/duplicates/groups
- /folders/tree photo_count subquery
- /tags count_subq (drives Tags + People sidebar counts)
- services/duplicates.regroup_duplicates (so hidden dupes never
contaminate the Duplicates view)
- services/search.hybrid_search — both semantic (pgvector) and FTS
legs join photos so rankings don't include hidden results
Intentionally NOT filtered:
- /photos?folder_id=X and /photos?heap_id=X (user-intentional browse)
- /library/maintenance/pipeline-stats (tracks real worker state)
- cleanup service (disk-level ops, not views)
Frontend:
- sourceFolders.setHidden(id, hidden) API client method
- FolderTreeNode.is_hidden carried through the tree into TreeItem
- LeftSidebar kebab menu: "Hide from views" / "Show in views" with a
mutation that invalidates folders, photos, stats, and tags caches
- Hidden folder rows swap the Folder icon for EyeOff and render the
label italic/muted so the state is visible at a glance
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Three overlapping fixes so the ingestion pipeline actually runs and the
user can see what it's doing:
Pipeline recovery
- app/database.py: use NullPool when MULITA_CELERY_WORKER=1 so each
Celery task opens a fresh asyncpg connection on its own event loop.
Fixes "another operation in progress" and "Future attached to a
different loop" errors that were dropping ~every thumbnail +
extract_metadata task on the floor.
- app/tasks/thumbs.py: initialize photo=None before the try and rollback
on error so a transport failure in the initial SELECT doesn't raise
UnboundLocalError in the except block and leak rows stuck in 'pending'.
- app/services/vision/bootstrap_models.py: on missing model files,
invoke export_models automatically instead of just warning. First
boot of a fresh install now self-heals.
- app/services/vision/export_models.py: shutil.move instead of
Path.rename so the YOLO export survives the /app → /data/models
cross-volume hop.
- requirements.txt: add ultralytics so export works in a stock image.
Worker topology
- docker-compose.yml: replace the single worker with worker-light
(default/high/low queues, c=2, IO-bound) and worker-vision (vision
queue, c=5, OMP_NUM_THREADS=1 to avoid oversubscription on 6 cores).
Vision is pinned to ≤5 parallel inferences so ONNX doesn't each
spawn an all-cores intra-op pool.
- .env / .env.example: CELERYD_CONCURRENCY replaced with
CELERY_LIGHT_CONCURRENCY + CELERY_VISION_CONCURRENCY.
- Backfill queries in thumbs / scan / vision now ORDER BY taken_at
DESC NULLS LAST so newest photos finish first — the library fills
in top-down in the UI instead of arbitrary insertion order.
Settings visibility
- routers/library.py: new GET /maintenance/pipeline-stats returning
done/total per stage (thumbnails, exif, gps, phash, embeddings,
tags, ocr, faces, face clusters, duplicate groups). Worker-status
now also reports the `vision` queue depth, which was missing.
- services/api.ts: PipelineStats / PipelineStage / ScanStatus types
and the matching client call.
- components/dialogs/SettingsDialog.tsx:
- new Pipeline Progress card with one progress bar per stage
- inline scan banner (processed/total/current folder) inside the
Library section while a scan is running
- Tasks/min throughput computed by diffing worker processed counters
between polls
- Workers section calls out the vision queue and documents the
CELERY_LIGHT/VISION_CONCURRENCY + docker compose up -d scale path
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
PyTorch default install pulls ~7GB of CUDA libs, exceeding disk on small
VMs. Switching to CPU-only saves ~6GB. Also run create_all before alembic
so migrations find existing tables on a fresh Postgres.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace Info icon with braille mule art in the right sidebar empty
state. Lighten text-muted (#a8997d → #c4b599) and text-faint
(#5e5448 → #7a6e5e) for better readability across the app.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Reflect current state: Postgres+pgvector replaces SQLite, vision
pipeline (YOLO, CLIP, InsightFace, OCR) is shipped, card-grid browse
views for tags/colors/ratings/people, map view, duplicate detection,
and semantic search are all live. Remove completed items from future
features.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace Timeline-based grouped views (tags, colors, rated) with
dedicated card-grid components that drill into Timeline detail views
on click/Enter. Adds shared useCardGridNav hook for arrow-key
navigation across all four card grids (tags, colors, rated, people).
- TagsView, ColorsView, RatedView: card grid → inline Timeline detail
- PeopleView: migrated to same pattern (Timeline replaces custom grid)
- Tags endpoint: fall back to first associated photo for representative
- Filter store: add ratingMax for exact rating filtering in RatedView
- Timeline: remove tag/rating/color grouping; skip date headers when
groupBy != 'date' so detail views render flat grids
- SettingsDialog: bump z-index above Leaflet map layers
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Cross-reference face detections with YOLO 'person' bounding boxes —
only keep faces that overlap >= 50% with a detected human body. This
eliminates false positives on dogs, paintings, and cartoons without
needing an aggressive score threshold.
Lower face detection threshold back to 0.6 since the person-overlap
check is now the primary precision filter.
Tested: 6 verified faces from 4 photos, zero false positives.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Eliminates false positives (dog faces, painting faces) by requiring
score >= 0.85. Tighten cluster eps to 0.25 for better separation.
Tested: 4 real faces from 2 photos, no false positives.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Fix Photo.created_at → Photo.added_at (column doesn't exist)
- Fix Postgres DISTINCT + ORDER BY conflict by using a subquery for
tag_id filtering instead of JOIN + DISTINCT on the outer query
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Clicking a person card now opens a detail sub-view within the People
section showing their photo grid. Back arrow returns to the card grid.
Photos are clickable to open the preview. Rename is available in both
the card grid and the detail header.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- recluster_faces now writes photo_tags rows for each face cluster so
the tag count and tag_ids filter work (previously count was always 0)
- Old cluster tags and photo_tags are cleaned up before re-clustering
- Raise face detection threshold from 0.4 to 0.7 to reduce false
positives (was detecting dog faces as people)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Rewrite faces.py to use cv2.FaceDetectorYN instead of raw ONNX
(handles multi-scale anchor decoding and NMS internally)
- Load original photo files at up to 4000px for face detection instead
of 240px thumbnails — faces were too small to detect at thumbnail res
- Falls back to thumbnail if original is unavailable
Tested: 33 faces extracted from 13 photos, clustered into 1 person.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
