migrate api from chroma to sqlite_vec

frigate/api/event.py

@@ -1,8 +1,6 @@
 """Event apis."""
 
-import base64
 import datetime
-import io
 import logging
 import os
 from functools import reduce
@@ -10,12 +8,10 @@ from pathlib import Path
 from urllib.parse import unquote
 
 import cv2
-import numpy as np
 from fastapi import APIRouter, Request
 from fastapi.params import Depends
 from fastapi.responses import JSONResponse
 from peewee import JOIN, DoesNotExist, fn, operator
-from PIL import Image
 from playhouse.shortcuts import model_to_dict
 
 from frigate.api.defs.events_body import (
@@ -39,7 +35,6 @@ from frigate.const import (
     CLIPS_DIR,
 )
 from frigate.embeddings import EmbeddingsContext
-from frigate.embeddings.embeddings import get_metadata
 from frigate.models import Event, ReviewSegment, Timeline
 from frigate.object_processing import TrackedObject
 from frigate.util.builtin import get_tz_modifiers
@@ -389,6 +384,8 @@ def events_search(request: Request, params: EventsSearchQueryParams = Depends())
 
     context: EmbeddingsContext = request.app.embeddings
 
+    logger.info(f"context: {context.embeddings}, conn: {context.embeddings.conn}")
+
     selected_columns = [
         Event.id,
         Event.camera,
@@ -484,14 +481,10 @@ def events_search(request: Request, params: EventsSearchQueryParams = Depends())
     else:
         event_ids = []
 
-    # Build the Chroma where clause based on the event IDs
+    thumb_results = []
-    where = {"id": {"$in": event_ids}} if event_ids else {}
+    desc_results = []
-
-    thumb_ids = {}
-    desc_ids = {}
 
     if search_type == "similarity":
-        # Grab the ids of events that match the thumbnail image embeddings
         try:
             search_event: Event = Event.get(Event.id == event_id)
         except DoesNotExist:
@@ -504,62 +497,25 @@ def events_search(request: Request, params: EventsSearchQueryParams = Depends())
                 ),
                 status_code=404,
             )
-        thumbnail = base64.b64decode(search_event.thumbnail)
+        thumb_results = context.embeddings.search_thumbnail(search_event.id, limit)
-        img = np.array(Image.open(io.BytesIO(thumbnail)).convert("RGB"))
-        thumb_result = context.embeddings.thumbnail.query(
-            query_images=[img],
-            n_results=limit,
-            where=where,
-        )
-        thumb_ids = dict(
-            zip(
-                thumb_result["ids"][0],
-                context.thumb_stats.normalize(thumb_result["distances"][0]),
-            )
-        )
     else:
         search_types = search_type.split(",")
 
         if "thumbnail" in search_types:
-            thumb_result = context.embeddings.thumbnail.query(
+            thumb_results = context.embeddings.search_thumbnail(query, limit)
-                query_texts=[query],
-                n_results=limit,
-                where=where,
-            )
-            # Do a rudimentary normalization of the difference in distances returned by CLIP and MiniLM.
-            thumb_ids = dict(
-                zip(
-                    thumb_result["ids"][0],
-                    context.thumb_stats.normalize(thumb_result["distances"][0]),
-                )
-            )
 
         if "description" in search_types:
-            desc_result = context.embeddings.description.query(
+            desc_results = context.embeddings.search_description(query, limit)
-                query_texts=[query],
-                n_results=limit,
-                where=where,
-            )
-            desc_ids = dict(
-                zip(
-                    desc_result["ids"][0],
-                    context.desc_stats.normalize(desc_result["distances"][0]),
-                )
-            )
 
     results = {}
-    for event_id in thumb_ids.keys() | desc_ids:
+    for result in thumb_results + desc_results:
-        min_distance = min(
+        event_id, distance = result[0], result[1]
-            i
+        if event_id in event_ids or not event_ids:
-            for i in (thumb_ids.get(event_id), desc_ids.get(event_id))
+            if event_id not in results or distance < results[event_id]["distance"]:
-            if i is not None
+                results[event_id] = {
-        )
+                    "distance": distance,
-        results[event_id] = {
+                    "source": "thumbnail" if result in thumb_results else "description",
-            "distance": min_distance,
+                }
-            "source": "thumbnail"
-            if min_distance == thumb_ids.get(event_id)
-            else "description",
-        }
 
     if not results:
         return JSONResponse(content=[])
@@ -975,10 +931,9 @@ def set_description(
     # If semantic search is enabled, update the index
     if request.app.frigate_config.semantic_search.enabled:
         context: EmbeddingsContext = request.app.embeddings
-        context.embeddings.description.upsert(
+        context.embeddings.upsert_description(
-            documents=[new_description],
+            event_id=event_id,
-            metadatas=[get_metadata(event)],
+            description=new_description,
-            ids=[event_id],
         )
 
     response_message = (
@@ -1065,8 +1020,8 @@ def delete_event(request: Request, event_id: str):
     # If semantic search is enabled, update the index
     if request.app.frigate_config.semantic_search.enabled:
         context: EmbeddingsContext = request.app.embeddings
-        context.embeddings.thumbnail.delete(ids=[event_id])
+        context.embeddings.delete_thumbnail(id=[event_id])
-        context.embeddings.description.delete(ids=[event_id])
+        context.embeddings.delete_description(id=[event_id])
     return JSONResponse(
         content=({"success": True, "message": "Event " + event_id + " deleted"}),
         status_code=200,