From 890ea6c8d16f683140685b1c6300710147b7be32 Mon Sep 17 00:00:00 2001
From: Manav Rathi <manav@mrmr.io>
Date: Fri, 9 Aug 2024 12:22:55 +0530
Subject: [PATCH] Closer

---
 web/apps/photos/src/services/searchService.ts |  4 ++-
 .../new/photos/services/ml/cluster-new.ts     | 32 +++++++++++--------
 web/packages/new/photos/services/ml/index.ts  |  4 +--
 3 files changed, 24 insertions(+), 16 deletions(-)

diff --git a/web/apps/photos/src/services/searchService.ts b/web/apps/photos/src/services/searchService.ts
index 8488462aba..c697c34575 100644
--- a/web/apps/photos/src/services/searchService.ts
+++ b/web/apps/photos/src/services/searchService.ts
@@ -421,7 +421,9 @@ async function getAllPeople(limit: number = undefined) {
     //     people.push(person);
     // });
     people = people ?? [];
-    return people
+    const result = people
         .sort((p1, p2) => p2.files.length - p1.files.length)
         .slice(0, limit);
+    // log.debug(() => ["getAllPeople", result]);
+    return result;
 }
diff --git a/web/packages/new/photos/services/ml/cluster-new.ts b/web/packages/new/photos/services/ml/cluster-new.ts
index bee7c2fee8..3a615471d4 100644
--- a/web/packages/new/photos/services/ml/cluster-new.ts
+++ b/web/packages/new/photos/services/ml/cluster-new.ts
@@ -81,34 +81,40 @@ export const clusterFaces = (faceIndexes: FaceIndex[]) => {
     const clusters: FaceCluster[] = [];
     const clusterIndexByFaceID = new Map<string, number>();
     for (const [i, { faceID, embedding }] of faces.entries()) {
-        let j = 0;
-        for (; j < i; j++) {
-            // Can't find a better way for avoiding the null assertion.
+        // Find the nearest neighbour from among the faces we have already seen.
+        let nnIndex: number | undefined;
+        let nnCosineSimilarity = 0;
+        for (let j = 0; j < i; j++) {
+            // Can't find a way of avoiding the null assertion.
             // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
             const n = faces[j]!;
 
-            // TODO-ML: The distance metric and the thresholds are placeholders.
-
             // The vectors are already normalized, so we can directly use their
             // dot product as their cosine similarity.
             const csim = dotProduct(embedding, n.embedding);
-            if (csim > 0.5) {
-                // Found a neighbour near enough. Add this face to the
-                // neighbour's cluster and call it a day.
-                const ci = ensure(clusterIndexByFaceID.get(n.faceID));
-                clusters[ci]?.faceIDs.push(faceID);
-                clusterIndexByFaceID.set(faceID, ci);
-                break;
+            if (csim > 0.76 && csim > nnCosineSimilarity) {
+                nnIndex = j;
+                nnCosineSimilarity = csim;
             }
         }
-        if (j == i) {
+        if (nnIndex === undefined) {
             // We didn't find a neighbour. Create a new cluster with this face.
+
             const cluster = {
                 id: newNonSecureID("cluster_"),
                 faceIDs: [faceID],
             };
             clusters.push(cluster);
             clusterIndexByFaceID.set(faceID, clusters.length);
+        } else {
+            // Found a neighbour near enough. Add this face to the neighbour's
+            // cluster.
+
+            // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+            const nn = faces[nnIndex]!;
+            const nnClusterIndex = ensure(clusterIndexByFaceID.get(nn.faceID));
+            clusters[nnClusterIndex]?.faceIDs.push(faceID);
+            clusterIndexByFaceID.set(faceID, nnClusterIndex);
         }
     }
 
diff --git a/web/packages/new/photos/services/ml/index.ts b/web/packages/new/photos/services/ml/index.ts
index 434568e22f..bee9291563 100644
--- a/web/packages/new/photos/services/ml/index.ts
+++ b/web/packages/new/photos/services/ml/index.ts
@@ -309,7 +309,7 @@ export const indexNewUpload = (enteFile: EnteFile, uploadItem: UploadItem) => {
     void worker().then((w) => w.onUpload(enteFile, uploadItem));
 };
 
-let last: Person[] = [];
+let last: Person[] | undefined;
 
 /**
  * WIP! Don't enable, dragon eggs are hatching here.
@@ -318,7 +318,7 @@ export const wipCluster = async () => {
     if (!isDevBuild || !(await isInternalUser())) return;
     if (!process.env.NEXT_PUBLIC_ENTE_WIP_CL) return;
 
-    if (last.length) return last;
+    if (last) return last;
 
     const clusters = clusterFaces(await faceIndexes());