OCA\FaceRecognition\BackgroundJob\Tasks\CreateClustersTask::createClusterIfNeeded()

Complexity

Conditions	10
Paths	10

Size

Total Lines	82
Code Lines	43

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	0
CRAP Score	110

Importance

Changes

0

<?php
	private function createClusterIfNeeded(string $userId) {
		// Check that we processed enough images to start creating clusters
		//
		$modelId = intval($this->config->getAppValue('facerecognition', 'model', AddDefaultFaceModel::DEFAULT_FACE_MODEL_ID));

		$hasPersons = $this->personMapper->countPersons($userId) > 0;

		// Depending on whether we already have clusters, decide if we should create/recreate them.
		//
		if ($hasPersons) {
			// OK, we already got some persons. We now need to evaluate whether we want to recreate clusters.
			// We want to recreate clusters/persons if:
			// * Some cluster/person is invalidated (is_valid is false for someone)
			//     This means some image that belonged to this user is changed, deleted etc.
			// * There are some new faces. Now, we don't want to jump the gun here. We want to either have:
			// ** more than 10 new faces, or
			// ** less than 10 new faces, but they are older than 2h
			//  (basically, we want to avoid recreating cluster for each new face being uploaded,
			//  however, we don't want to wait too much as clusters could be changed a lot)
			//
			$haveNewFaces = false;
			$facesWithoutPersons = $this->faceMapper->countFaces($userId, $modelId, true);
			$this->logDebug(sprintf('Found %d faces without associated persons for user %s and model %d',
				$facesWithoutPersons, $userId, $modelId));
			// todo: get rid of magic numbers (move to config)
			if ($facesWithoutPersons >= 10) {
				$haveNewFaces = true;
			} else if ($facesWithoutPersons > 0) {
				// We have some faces, but not that many, let's see when oldest one is generated.
				$face = $this->faceMapper->getOldestCreatedFaceWithoutPerson($userId, $modelId);
				$oldestFaceTimestamp = $face->creationTime->getTimestamp();
				$currentTimestamp = (new \DateTime())->getTimestamp();
				$this->logDebug(sprintf('Oldest face without persons for user %s and model %d is from %s',
					$userId, $modelId, $face->creationTime->format('Y-m-d H:i:s')));
				// todo: get rid of magic numbers (move to config)
				if ($currentTimestamp - $oldestFaceTimestamp > 2 * 60 * 60) {
					$haveNewFaces = true;
				}
			}

			$stalePersonsCount = $this->personMapper->countPersons($userId, true);
			$this->logDebug(sprintf('Found %d stale persons for user %s and model %d', $stalePersonsCount, $userId, $modelId));
			$haveStalePersons = $stalePersonsCount > 0;

			if ($haveStalePersons === false && $haveNewFaces === false) {
				// If there is no invalid persons, and there is no recent new faces, no need to recreate cluster
				$this->logInfo('Clusters already exist, calculated there is no need to recreate them');
				return;
			}
		} else {
			// These are basic criteria without which we should not even consider creating clusters.
			// These clusters will be small and not "stable" enough and we should better wait for more images to come.
			// todo: 2 queries to get these 2 counts, can we do this smarter?
			$imageCount = $this->imageMapper->countUserImages($userId, $modelId);
			$imageProcessed = $this->imageMapper->countUserProcessedImages($userId, $modelId);
			$percentImagesProcessed = $imageProcessed / floatval($imageCount);
			$facesCount = $this->faceMapper->countFaces($userId, $modelId);
			// todo: get rid of magic numbers (move to config)
			if (($facesCount < 1000) && ($imageCount < 100) && ($percentImagesProcessed < 0.95)) {
				$this->logInfo(
					'Skipping cluster creation, not enough data (yet) collected. ' .
					'For cluster creation, you need either one of the following:');
				$this->logInfo(sprintf('* have 1000 faces already processed (you have %d),', $facesCount));
				$this->logInfo(sprintf('* have 100 images (you have %d),', $imageCount));
				$this->logInfo(sprintf('* or you need to have 95%% of you images processed (you have %.2f%%)', $percentImagesProcessed));
				return;
			}
		}

		$faces = $this->faceMapper->getFaces($userId, $modelId);
		$this->logInfo(count($faces) . ' faces found for clustering');

		// Cluster is associative array where key is person ID.
		// Value is array of face IDs. For old clusters, person IDs are some existing person IDs,
		// and for new clusters is whatever chinese whispers decides to identify them.
		//
		$currentClusters = $this->getCurrentClusters($faces);
		$newClusters = $this->getNewClusters($faces);
		$this->logInfo(count($newClusters) . ' clusters found for clustering');
		// New merge
		$mergedClusters = $this->mergeClusters($currentClusters, $newClusters);
		$this->personMapper->mergeClusterToDatabase($userId, $currentClusters, $mergedClusters);
	}