endoreg-db 0.8.2.7__py3-none-any.whl → 0.8.2.9__py3-none-any.whl

endoreg_db/models/metadata/model_meta.py

@@ -128,6 +128,20 @@ class ModelMeta(models.Model):
             cls, meta_name, model_name, labelset_name, weights_file,
             requested_version, bump_if_exists, **kwargs
         )
+        
+    @classmethod
+    def setup_default_from_huggingface(
+        cls: Type["ModelMeta"],
+        model_id: str,
+        task: str = "image-classification",
+        labelset_name: Optional[str] = None,
+    ) -> "ModelMeta":
+        """
+        Downloads a pretrained model from Hugging Face and initializes ModelMeta automatically.
+        """
+        from . import model_meta_logic as logic
+        return logic.setup_default_from_huggingface_logic(cls, model_id, task, labelset_name)
+
 
     @classmethod
     def get_latest_version_number(cls: Type["ModelMeta"], meta_name: str, model_name: str) -> int:

endoreg_db/models/metadata/model_meta_logic.py

@@ -1,7 +1,7 @@
 import shutil
 from pathlib import Path
 from typing import Optional, TYPE_CHECKING, Any, Type
-
+from huggingface_hub import hf_hub_download
 from django.db import transaction
 
 # Assuming ModelMeta, AiModel, LabelSet are importable from the correct locations
@@ -234,3 +234,90 @@ def get_model_meta_by_name_version_logic(
             raise cls.DoesNotExist(
                 f"No ModelMeta found for '{meta_name}' and model '{model_name}'."
             )
+            
+from huggingface_hub import model_info
+import re
+
+def infer_default_model_meta_from_hf(model_id: str) -> dict[str, Any]:
+    """
+    Infers default model metadata (activation, normalization, input size)
+    from a Hugging Face model_id using its tags and architecture.
+
+    Returns:
+        A dict with fields: name, activation, mean, std, size_x, size_y
+    """
+
+    if not (info := model_info(model_id)):
+        logger.info(f"Could not retrieve model info for {model_id}, using ColoReg segmentation defaults.")
+        return {
+            "name": "wg-lux/colo_segmentation_RegNetX800MF_base",
+            "activation": "sigmoid",
+            "mean": (0.45211223, 0.27139644, 0.19264949),
+            "std": (0.31418097, 0.21088019, 0.16059452),
+            "size_x": 716,
+            "size_y": 716,
+            "description": f"Defaults for unknown model {model_id}",
+        }
+
+    # Extract architecture from tags or model_id ---
+    tags = info.tags or []
+    model_name = model_id.split("/")[-1].lower()
+
+    # Heuristics for architecture and task
+    architecture = next((t for t in tags if t.startswith("architecture:")), None)
+    task = next((t for t in tags if t.startswith("task:")), None)
+
+    # Default values
+    activation = "sigmoid"
+    size_x = size_y = 716
+    mean = (0.45211223, 0.27139644, 0.19264949)
+    std = (0.31418097, 0.21088019, 0.16059452)
+
+    # --- 2. Task-based inference ---
+    if task:
+        if "segmentation" in task or "detection" in task:
+            activation = "sigmoid"
+        elif any(k in task for k in ["classification"]):
+            activation = "softmax"
+
+    # --- 3. Architecture-based inference ---
+    if architecture:
+        arch = architecture.replace("architecture:", "")
+    else:
+        arch = re.sub(r"[^a-z0-9]+", "_", model_name)
+
+    return {
+        "name": arch,
+        "activation": activation,
+        "mean": mean,
+        "std": std,
+        "size_x": size_x,
+        "size_y": size_y,
+        "description": f"Inferred defaults for {model_id}",
+    }
+    
+def setup_default_from_huggingface_logic(cls, model_id: str, labelset_name: str | None = None):
+    """
+    Downloads model weights from Hugging Face and auto-fills ModelMeta fields.
+    """
+    meta = infer_default_model_meta_from_hf(model_id)
+
+    # Download weights
+    weights_path = hf_hub_download(repo_id=model_id, filename="pytorch_model.bin", local_dir=WEIGHTS_DIR)
+
+    ai_model, _ = AiModel.objects.get_or_create(name=meta["name"])
+    labelset = LabelSet.objects.first() if not labelset_name else LabelSet.objects.get(name=labelset_name)
+
+    return create_from_file_logic(
+        cls,
+        meta_name=meta["name"],
+        model_name=ai_model.name,
+        labelset_name=labelset.name,
+        weights_file=weights_path,
+        activation=meta["activation"],
+        mean=meta["mean"],
+        std=meta["std"],
+        size_x=meta["size_x"],
+        size_y=meta["size_y"],
+        description=meta["description"],
+    )

endoreg_db/services/video_import.py

@@ -523,20 +523,7 @@ class VideoImportService():
             video = self.current_video
             if video is None:
                 self.logger.warning("No VideoFile instance available for fallback anonymization")
-            else:
-                # Try VideoFile.pipe_2() method if available
-                if hasattr(video, 'pipe_2'):
-                    self.logger.info("Trying VideoFile.pipe_2() method...")
-                    if video.pipe_2():
-                        self.logger.info("VideoFile.pipe_2() succeeded")
-                        self.processing_context['anonymization_completed'] = True
-                        return
-                    self.logger.warning("VideoFile.pipe_2() returned False")
-                # Try direct anonymization via _anonymize
-                if _anonymize(video, delete_original_raw=self.delete_source):
-                    self.logger.info("VideoFile._anonymize() succeeded")
-                    self.processing_context['anonymization_completed'] = True
-                    return
+
 
             # Strategy 2: Simple copy (no processing, just copy raw to processed)
             self.logger.info("Using simple copy fallback (raw video will be used as 'processed' video)")
@@ -888,10 +875,8 @@ class VideoImportService():
         video_filename = self.processing_context.get('video_filename', Path(raw_video_path).name)
         cleaned_filename = f"cleaned_{video_filename}"
         cleaned_video_path = Path(raw_video_path).parent / cleaned_filename
-        
-        processor_roi, endoscope_roi = self._get_processor_roi_info(video)
-        
-        # Processor roi can be used later to OCR preknown regions.
+                
+        # Processor roi is used later to OCR preknown regions.
         
         # Clean video with ROI masking (heavy I/O operation)
         actual_cleaned_path, extracted_metadata = frame_cleaner.clean_video(
@@ -903,29 +888,6 @@ class VideoImportService():
             technique="mask_overlay"
         )
         
-        # Optional: enrich metadata using TrOCR+LLM on one random extracted frame
-        try:
-            # Prefer frames belonging to this video (UUID in path), else pick any frame
-            frame_candidates = list(RAW_FRAME_DIR.rglob("*.jpg")) + list(RAW_FRAME_DIR.rglob("*.png"))
-            video_uuid = str(video.uuid)
-            filtered = [p for p in frame_candidates if video_uuid in str(p)] or frame_candidates
-            if filtered:
-                sample_frame = random.choice(filtered)
-                ocr_text = trocr_full_image_ocr(sample_frame)
-                if ocr_text:
-                    llm_metadata = frame_cleaner.extract_metadata(ocr_text)
-                    if llm_metadata:
-                        # Merge with already extracted frame-level metadata
-                        extracted_metadata = frame_cleaner.frame_metadata_extractor.merge_metadata(
-                            extracted_metadata or {}, llm_metadata
-                        )
-                        self.logger.info("LLM metadata extraction (random frame) successful")
-                    else:
-                        self.logger.info("LLM metadata extraction (random frame) found no data")
-                else:
-                    self.logger.info("No text extracted by TrOCR on random frame")
-        except Exception as e:
-            self.logger.error(f"LLM metadata enrichment step failed: {e}")
         
         # Store cleaned video path for later use in _cleanup_and_archive
         self.processing_context['cleaned_video_path'] = actual_cleaned_path
@@ -1048,8 +1010,6 @@ class VideoImportService():
                     self.processed_files.remove(file_path_str)
                     self.logger.info(f"Removed {file_path_str} from processed files (failed processing)")
             
-                 
-            
             
         except Exception as e:
             self.logger.warning(f"Error during context cleanup: {e}")

{endoreg_db-0.8.2.7.dist-info → endoreg_db-0.8.2.9.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: endoreg-db
-Version: 0.8.2.7
+Version: 0.8.2.9
 Summary: EndoReg Db Django App
 Project-URL: Homepage, https://info.coloreg.de
 Project-URL: Repository, https://github.com/wg-lux/endoreg-db
@@ -29,10 +29,11 @@ Requires-Dist: dotenv>=0.9.9
 Requires-Dist: faker>=37.6.0
 Requires-Dist: flake8>=7.3.0
 Requires-Dist: gunicorn>=23.0.0
+Requires-Dist: huggingface-hub>=0.35.3
 Requires-Dist: icecream>=2.1.4
 Requires-Dist: librosa==0.11.0
 Requires-Dist: llvmlite>=0.44.0
-Requires-Dist: lx-anonymizer[llm,ocr]>=0.8.5
+Requires-Dist: lx-anonymizer[llm,ocr]>=0.8.7
 Requires-Dist: moviepy==2.2.1
 Requires-Dist: mypy>=1.16.0
 Requires-Dist: numpy>=2.2.3

{endoreg_db-0.8.2.7.dist-info → endoreg_db-0.8.2.9.dist-info}/RECORD

@@ -461,8 +461,8 @@ endoreg_db/models/medical/risk/risk.py,sha256=g5pgAfCfsvH88nbmX3xsASF3OZgNA-G6NJ
 endoreg_db/models/medical/risk/risk_type.py,sha256=kEugcaWSTEWH_Vxq4dcF80Iv1L4_Kk1JKJGQMgz_s0o,1350
 endoreg_db/models/metadata/__init__.py,sha256=8I6oLj3YTmeaPGJpL0AWG5gLwp38QzrEggxSkTisv7c,474
 endoreg_db/models/metadata/frame_ocr_result.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-endoreg_db/models/metadata/model_meta.py,sha256=aZH6Bz5Ss874Knvg1b3Kgq6gU8kVzPHXneunZJNF4yw,7111
-endoreg_db/models/metadata/model_meta_logic.py,sha256=yiIWbxxykUp6VB_7imRqSXcO0RS5GuoYP83O48TyKws,8987
+endoreg_db/models/metadata/model_meta.py,sha256=AveneX6f4AOQD4R-IQB8UqcqTzKpek2yuEWk7ybtF38,7606
+endoreg_db/models/metadata/model_meta_logic.py,sha256=27mqScxUTJXNUVc6CqAs5dXjspEsh0TWPmlxdJVulGc,12015
 endoreg_db/models/metadata/pdf_meta.py,sha256=BTmpSgqxmPKi0apcNjyrZAS4AFKCPXVdBd6VBeyyv6E,3174
 endoreg_db/models/metadata/sensitive_meta.py,sha256=ekLHrW-b5uYcjfkRd0EW5ncx5ef8Bu-K6msDkpWCAbk,13034
 endoreg_db/models/metadata/sensitive_meta_logic.py,sha256=Oh7ssZQEPfKGfRMF5nXKJpOIxXx-Xibd3rpOu-bQilk,29988
@@ -600,7 +600,7 @@ endoreg_db/services/pseudonym_service.py,sha256=CJhbtRa6K6SPbphgCZgEMi8AFQtB18CU
 endoreg_db/services/requirements_object.py,sha256=290zf8AEbVtCoHhW4Jr7_ud-RvrqYmb1Nz9UBHtTnc0,6164
 endoreg_db/services/segment_sync.py,sha256=YgHvIHkbW4mqCu0ACf3zjRSZnNfxWwt4gh5syUVXuE0,6400
 endoreg_db/services/storage_aware_video_processor.py,sha256=kKFK64vXLeBSVkp1YJonU3gFDTeXZ8C4qb9QZZB99SE,13420
-endoreg_db/services/video_import.py,sha256=X20FQkEO5QGcfuacAz8jX1_LW1GhwbF33JGPpmypEyk,50161
+endoreg_db/services/video_import.py,sha256=PhcOgxU5M4uSEklBXEWHpIaNX-yIYv1rJy-T-fCU8cs,47830
 endoreg_db/tasks/upload_tasks.py,sha256=OJq7DhNwcbWdXzHY8jz5c51BCVkPN5gSWOz-6Fx6W5M,7799
 endoreg_db/tasks/video_ingest.py,sha256=kxFuYkHijINV0VabQKCFVpJRv6eCAw07tviONurDgg8,5265
 endoreg_db/tasks/video_processing_tasks.py,sha256=KjcERRJ1TZzmavBpvr6OsvSTUViU0PR1ECWnEdzu2Js,14140
@@ -784,7 +784,7 @@ endoreg_db/views/video/video_meta.py,sha256=C1wBMTtQb_yzEUrhFGAy2UHEWMk_CbU75WXX
 endoreg_db/views/video/video_processing_history.py,sha256=mhFuS8RG5GV8E-lTtuD0qrq-bIpnUFp8vy9aERfC-J8,770
 endoreg_db/views/video/video_remove_frames.py,sha256=2FmvNrSPM0fUXiBxINN6vBUUDCqDlBkNcGR3WsLDgKo,1696
 endoreg_db/views/video/video_stream.py,sha256=kLyuf0ORTmsLeYUQkTQ6iRYqlIQozWhMMR3Lhfe_trk,12148
-endoreg_db-0.8.2.7.dist-info/METADATA,sha256=BCQcbq7ZExYZF5gBfmkDflNtysGjAt36WUMKmqfA-48,14719
-endoreg_db-0.8.2.7.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-endoreg_db-0.8.2.7.dist-info/licenses/LICENSE,sha256=OXLcl0T2SZ8Pmy2_dmlvKuetivmyPd5m1q-Gyd-zaYY,35149
-endoreg_db-0.8.2.7.dist-info/RECORD,,
+endoreg_db-0.8.2.9.dist-info/METADATA,sha256=2PdVUGEhQwGP1CvM1194ox8-B7Ir5wG70n0iwnZdJQ0,14758
+endoreg_db-0.8.2.9.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+endoreg_db-0.8.2.9.dist-info/licenses/LICENSE,sha256=OXLcl0T2SZ8Pmy2_dmlvKuetivmyPd5m1q-Gyd-zaYY,35149
+endoreg_db-0.8.2.9.dist-info/RECORD,,