diff --git a/mobile/lib/face/db.dart b/mobile/lib/face/db.dart
index c8b4f75f7..bc4b15885 100644
--- a/mobile/lib/face/db.dart
+++ b/mobile/lib/face/db.dart
@@ -154,7 +154,7 @@ class FaceMLDataDB {
     final Map<int, int> result = {};
     final db = await instance.database;
     final List<Map<String, dynamic>> maps = await db.rawQuery(
-      'SELECT $fileIDColumn, COUNT(*) as count FROM $facesTable where $faceScore > 0.8 GROUP BY $fileIDColumn',
+      'SELECT $fileIDColumn, COUNT(*) as count FROM $facesTable where $faceScore > $kMinFaceDetectionScore GROUP BY $fileIDColumn',
     );
 
     for (final map in maps) {
@@ -197,7 +197,7 @@ class FaceMLDataDB {
       final clusterIDs =
           cluterRows.map((e) => e[cluserIDColumn] as int).toList();
       final List<Map<String, dynamic>> faceMaps = await db.rawQuery(
-        'SELECT * FROM $facesTable where $faceClusterId IN (${clusterIDs.join(",")}) AND $fileIDColumn in (${fileId.join(",")}) AND $faceScore > 0.8 ORDER BY $faceScore DESC',
+        'SELECT * FROM $facesTable where $faceClusterId IN (${clusterIDs.join(",")}) AND $fileIDColumn in (${fileId.join(",")}) AND $faceScore > $kMinHighQualityFaceScore ORDER BY $faceScore DESC',
       );
       if (faceMaps.isNotEmpty) {
         if (avatarFileId != null) {
@@ -341,7 +341,7 @@ class FaceMLDataDB {
   ///
   /// Only selects faces with score greater than [minScore] and blur score greater than [minClarity]
   Future<Map<String, (int?, Uint8List)>> getFaceEmbeddingMap({
-    double minScore = kMinFaceScore,
+    double minScore = kMinHighQualityFaceScore,
     int minClarity = kLaplacianThreshold,
     int maxRows = 20000,
   }) async {
@@ -398,7 +398,7 @@ class FaceMLDataDB {
         facesTable,
         columns: [faceIDColumn, faceEmbeddingBlob],
         where:
-            '$faceScore > 0.8 AND $faceBlur > $kLaplacianThreshold AND $fileIDColumn IN (${fileIDs.join(",")})',
+            '$faceScore > $kMinHighQualityFaceScore AND $faceBlur > $kLaplacianThreshold AND $fileIDColumn IN (${fileIDs.join(",")})',
         limit: batchSize,
         offset: offset,
         orderBy: '$faceIDColumn DESC',
diff --git a/mobile/lib/face/model/face.dart b/mobile/lib/face/model/face.dart
index c29c03ae1..af151c2ec 100644
--- a/mobile/lib/face/model/face.dart
+++ b/mobile/lib/face/model/face.dart
@@ -11,7 +11,7 @@ class Face {
 
   bool get isBlurry => blur < kLaplacianThreshold;
 
-  bool get hasHighScore => score > kMinFaceScore;
+  bool get hasHighScore => score > kMinHighQualityFaceScore;
 
   bool get isHighQuality => (!isBlurry) && hasHighScore;
 
diff --git a/mobile/lib/services/face_ml/face_filtering/face_filtering_constants.dart b/mobile/lib/services/face_ml/face_filtering/face_filtering_constants.dart
index 478161f22..a1970fd4d 100644
--- a/mobile/lib/services/face_ml/face_filtering/face_filtering_constants.dart
+++ b/mobile/lib/services/face_ml/face_filtering/face_filtering_constants.dart
@@ -1,8 +1,13 @@
+import "package:photos/services/face_ml/face_detection/yolov5face/onnx_face_detection.dart";
+
 /// Blur detection threshold
 const kLaplacianThreshold = 15;
 
 /// Default blur value
 const kLapacianDefault = 10000.0;
 
-/// The minimum score for a face to be considered a face
-const kMinFaceScore = 0.78;
\ No newline at end of file
+/// The minimum score for a face to be considered a high quality face for clustering and person detection
+const kMinHighQualityFaceScore = 0.78;
+
+/// The minimum score for a face to be detected, regardless of quality. Use [kMinHighQualityFaceScore] for high quality faces.
+const kMinFaceDetectionScore = YoloOnnxFaceDetection.kMinScoreSigmoidThreshold;
diff --git a/mobile/lib/services/face_ml/face_ml_service.dart b/mobile/lib/services/face_ml/face_ml_service.dart
index db3931ed7..a3b36e192 100644
--- a/mobile/lib/services/face_ml/face_ml_service.dart
+++ b/mobile/lib/services/face_ml/face_ml_service.dart
@@ -362,7 +362,9 @@ class FaceMlService {
     await clusterAllImages();
   }
 
-  Future<void> clusterAllImages({double minFaceScore = kMinFaceScore}) async {
+  Future<void> clusterAllImages({
+    double minFaceScore = kMinHighQualityFaceScore,
+  }) async {
     _logger.info("`clusterAllImages()` called");
 
     try {