dtoro/mule-image
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: vision tasks inherit user_id, admin owns mount root

Browse Source 
- 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>
This commit is contained in:
authentik Default Admin
2026-04-13 00:01:28 +02:00
parent 35d87a2749
commit fbeefb24a0
9 changed files with 35 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
backend/app/config.py
View File

@@ -33,7 +33,7 @@ class PerformanceSettings(BaseModel):
class EmbedderSettings(BaseModel):
"""CLIP / SigLIP embedding model settings.
Supported: "openclip_vitb32" (512-d), "siglip2_vitb16" (768-d, default)."""
name: str = "siglip2_vitb16"
name: str = "openclip_vitb32"
batch_size: int = 8
class OCRSettings(BaseModel):

2
backend/app/models/embeddings.py
View File

@@ -20,5 +20,5 @@ class Embedding(Base):
photo_id = Column(String, ForeignKey('photos.id', ondelete='CASCADE'), primary_key=True)
model = Column(String(64), primary_key=True) # e.g. 'siglip2_vitb16'
vector = Column(Vector(768)) # SigLIP2 ViT-B/16 → 768-d
vector = Column(Vector(512)) # OpenCLIP ViT-B/32 → 512-d
created_at = Column(DateTime(timezone=True), server_default=func.now())

6
backend/app/routers/auth.py
View File

@@ -161,7 +161,9 @@ async def setup(body: SetupRequest, db: AsyncSession = Depends(get_db)):
if len(body.password) < 6:
raise HTTPException(status_code=400, detail="Password must be at least 6 characters")
media_path = os.path.join(settings.photo_dirs, body.username.strip())
# The initial admin owns the entire photo mount root. Regular users
# (created later via admin panel) get a subdirectory under it.
media_path = settings.photo_dirs
os.makedirs(media_path, exist_ok=True)
user = User(
@@ -171,8 +173,8 @@ async def setup(body: SetupRequest, db: AsyncSession = Depends(get_db)):
media_path=media_path,
)
db.add(user)
await db.flush() # get user.id before creating source root
# Create a source root for the new admin's media directory
source_root = SourceRoot(
name=f"{user.username}'s Library",
path=media_path,

2
backend/app/services/vision/bootstrap_models.py
View File

@@ -29,8 +29,6 @@ DOWNLOADS = []
EXPORTS = [
("embed/visual.onnx", "OpenCLIP ViT-B/32 visual encoder"),
("embed/textual.onnx", "OpenCLIP ViT-B/32 textual encoder"),
("embed_siglip2/visual.onnx", "SigLIP2 ViT-B/16 visual encoder"),
("embed_siglip2/textual.onnx", "SigLIP2 ViT-B/16 textual encoder"),
("detect/yolov8n.onnx", "YOLOv8n object detector"),
]

2
backend/app/services/vision/classify.py
View File

@@ -63,7 +63,7 @@ class CLIPContentClassifier(ContentClassifier):
# image embeddings.
embedder_name = settings.embedder.name
if embedder_name.startswith("siglip2"):
model_arch = "ViT-B-16-SigLIP2"
model_arch = "ViT-B-16-SigLIP-384"
pretrained = "webli"
else:
model_arch = "ViT-B-32"

2
backend/app/services/vision/embed.py
View File

@@ -129,7 +129,7 @@ class SigLIP2Embedder(Embedder):
def embed_text(self, text: str) -> np.ndarray:
import open_clip
tokenizer = open_clip.get_tokenizer("ViT-B-16-SigLIP2")
tokenizer = open_clip.get_tokenizer("ViT-B-16-SigLIP-384")
tokens = tokenizer([text]).numpy().astype(np.int64)
inputs = self._textual.get_inputs()
feed = {inputs[0].name: tokens}

6
backend/app/services/vision/export_models.py
View File

@@ -135,9 +135,9 @@ def export_siglip2(models_dir: Path):
logger.info("SigLIP2 ONNX files already exist, skipping export")
return
logger.info("Loading SigLIP2 ViT-B-16-SigLIP2 webli...")
logger.info("Loading SigLIP2 ViT-B-16-SigLIP-384 webli...")
model, _, preprocess = open_clip.create_model_and_transforms(
"ViT-B-16-SigLIP2", pretrained="webli"
"ViT-B-16-SigLIP-384", pretrained="webli"
)
model.eval()
@@ -162,7 +162,7 @@ def export_siglip2(models_dir: Path):
# ── Textual encoder ───────────────────────────────────────────────
if not textual_path.exists():
logger.info("Exporting SigLIP2 textual encoder → %s", textual_path)
tokenizer = open_clip.get_tokenizer("ViT-B-16-SigLIP2")
tokenizer = open_clip.get_tokenizer("ViT-B-16-SigLIP-384")
dummy_text = tokenizer(["a photo"]).to(torch.int64)
class SigLIP2TextEncoder(torch.nn.Module):

29
backend/app/tasks/vision.py
View File

@@ -163,6 +163,12 @@ def detect_objects(photo_id: str):
session = _get_sync_session()
try:
# Get the photo's user_id so tags inherit ownership.
photo = session.execute(
select(Photo).where(Photo.id == photo_id)
).scalar_one_or_none()
owner_id = photo.user_id if photo else None
# Wipe previous detection results for this photo from this model
session.execute(
delete(photo_tags).where(
@@ -178,13 +184,13 @@ def detect_objects(photo_id: str):
best_per_label[det.label] = (det.confidence, det.bbox)
for label, (confidence, bbox) in best_per_label.items():
# Find or create the object tag
# Find or create the object tag (scoped to user)
tag = session.execute(
select(Tag).where(Tag.name == label, Tag.kind == 'object')
select(Tag).where(Tag.name == label, Tag.kind == 'object', Tag.user_id == owner_id)
).scalar_one_or_none()
if not tag:
tag = Tag(name=label, kind='object', source=source_name)
tag = Tag(name=label, kind='object', source=source_name, user_id=owner_id)
session.add(tag)
session.flush() # get tag.id
@@ -234,6 +240,12 @@ def classify_content(photo_id: str):
session = _get_sync_session()
try:
# Get the photo's user_id so tags inherit ownership.
photo = session.execute(
select(Photo).where(Photo.id == photo_id)
).scalar_one_or_none()
owner_id = photo.user_id if photo else None
# Wipe previous classification for this photo
session.execute(
delete(photo_tags).where(
@@ -242,13 +254,13 @@ def classify_content(photo_id: str):
)
)
# Find or create content_type tag
# Find or create content_type tag (scoped to user)
tag = session.execute(
select(Tag).where(Tag.name == best.label, Tag.kind == 'content_type')
select(Tag).where(Tag.name == best.label, Tag.kind == 'content_type', Tag.user_id == owner_id)
).scalar_one_or_none()
if not tag:
tag = Tag(name=best.label, kind='content_type', source=source_name)
tag = Tag(name=best.label, kind='content_type', source=source_name, user_id=owner_id)
session.add(tag)
session.flush()
@@ -434,11 +446,16 @@ def recluster_faces():
if label not in cluster_tag_map:
cluster_name = f"Person {label + 1}"
# Inherit user_id from the representative photo.
rep_photo = session.execute(
select(Photo.user_id).where(Photo.id == face_rows[i].photo_id)
).scalar_one_or_none()
tag = Tag(
name=cluster_name,
kind='face_cluster',
source=source_name,
representative_photo_id=face_rows[i].photo_id,
user_id=rep_photo,
)
session.add(tag)
session.flush()

1
backend/requirements.txt
View File

@@ -37,6 +37,7 @@ watchfiles==0.21.0
# Vision pipeline (ONNX Runtime CPU inference)
onnxruntime==1.18.1
open-clip-torch==2.24.0 # tokenizer + export helper; inference via ONNX
transformers>=4.37.0 # HuggingFace tokenizer for SigLIP models
ultralytics==8.4.37 # YOLOv8n export helper; inference via ONNX
rapidocr-onnxruntime==1.3.22
scikit-learn==1.4.0 # DBSCAN for face clustering