matiasdelellis /
facerecognition
| Total Complexity | 61 |
| Total Lines | 355 |
| Duplicated Lines | 0 % |
| Coverage | 66.49% |
| Changes | 6 | ||
| Bugs | 0 | Features | 1 |
Complex classes like CreateClustersTask often do a lot of different things. To break such a class down, we need to identify a cohesive component within that class. A common approach to find such a component is to look for fields/methods that share the same prefixes, or suffixes.
Once you have determined the fields that belong together, you can apply the Extract Class refactoring. If the component makes sense as a sub-class, Extract Subclass is also a candidate, and is often faster.
While breaking up the class, it is a good idea to analyze how other classes use CreateClustersTask, and based on these observations, apply Extract Interface, too.
| 1 | <?php |
||
| 41 | class CreateClustersTask extends FaceRecognitionBackgroundTask { |
||
| 42 | /** @var PersonMapper Person mapper*/ |
||
| 43 | private $personMapper; |
||
| 44 | |||
| 45 | /** @var ImageMapper Image mapper*/ |
||
| 46 | private $imageMapper; |
||
| 47 | |||
| 48 | /** @var FaceMapper Face mapper*/ |
||
| 49 | private $faceMapper; |
||
| 50 | |||
| 51 | /** @var SettingsService Settings service*/ |
||
| 52 | private $settingsService; |
||
| 53 | |||
| 54 | /** |
||
| 55 | * @param PersonMapper |
||
| 56 | * @param ImageMapper |
||
| 57 | * @param FaceMapper |
||
| 58 | * @param SettingsService |
||
| 59 | */ |
||
| 60 | 3 | public function __construct(PersonMapper $personMapper, |
|
| 71 | 3 | } |
|
| 72 | |||
| 73 | /** |
||
| 74 | * @inheritdoc |
||
| 75 | */ |
||
| 76 | 1 | public function description() { |
|
| 77 | 1 | return "Create new persons or update existing persons"; |
|
| 78 | } |
||
| 79 | |||
| 80 | /** |
||
| 81 | * @inheritdoc |
||
| 82 | */ |
||
| 83 | 1 | public function execute(FaceRecognitionContext $context) { |
|
| 104 | } |
||
| 105 | |||
| 106 | 1 | private function createClusterIfNeeded(string $userId) { |
|
| 107 | 1 | $modelId = $this->settingsService->getCurrentFaceModel(); |
|
| 108 | |||
| 109 | // Depending on whether we already have clusters, decide if we should create/recreate them. |
||
| 110 | // |
||
| 111 | 1 | $hasPersons = $this->personMapper->countPersons($userId, $modelId) > 0; |
|
| 112 | 1 | if ($hasPersons) { |
|
| 113 | $forceRecreate = $this->needRecreateBySettings($userId); |
||
| 114 | $haveEnoughFaces = $this->hasNewFacesToRecreate($userId, $modelId); |
||
| 115 | $haveStaled = $this->hasStalePersonsToRecreate($userId, $modelId); |
||
| 116 | |||
| 117 | if ($forceRecreate) { |
||
| 118 | $this->logInfo('Clusters already exist, but there was some change that requires recreating the clusters'); |
||
| 119 | } |
||
| 120 | else if ($haveEnoughFaces || $haveStaled) { |
||
| 121 | $this->logInfo('Face clustering will be recreated with new information or changes'); |
||
| 122 | } |
||
| 123 | else { |
||
| 124 | // If there is no invalid persons, and there is no recent new faces, no need to recreate cluster |
||
| 125 | $this->logInfo('Clusters already exist, estimated there is no need to recreate them'); |
||
| 126 | return; |
||
| 127 | } |
||
| 128 | } |
||
| 129 | else { |
||
| 130 | // User should not be able to use this directly, used in tests |
||
| 131 | 1 | $forceTestCreation = $this->settingsService->_getForceCreateClusters($userId); |
|
| 132 | 1 | $needCreate = $this->needCreateFirstTime($userId, $modelId); |
|
| 133 | |||
| 134 | 1 | if ($forceTestCreation) { |
|
| 135 | 1 | $this->logInfo('Force the creation of clusters for testing'); |
|
| 136 | } |
||
| 137 | 1 | else if ($needCreate) { |
|
| 138 | $this->logInfo('Face clustering will be created for the first time.'); |
||
| 139 | } |
||
| 140 | else { |
||
| 141 | 1 | $this->logInfo( |
|
| 142 | 'Skipping cluster creation, not enough data (yet) collected. ' . |
||
| 143 | 1 | 'For cluster creation, you need either one of the following:'); |
|
| 144 | 1 | $this->logInfo('* have 1000 faces already processed'); |
|
| 145 | 1 | $this->logInfo('* or you need to have 95% of you images processed'); |
|
| 146 | 1 | $this->logInfo('Use stats command to track progress'); |
|
| 147 | 1 | return; |
|
| 148 | } |
||
| 149 | } |
||
| 150 | |||
| 151 | // Ok. If we are here, the clusters must be recreated. |
||
| 152 | // |
||
| 153 | 1 | $faces = $this->faceMapper->getFaces($userId, $modelId); |
|
| 154 | 1 | $this->logInfo(count($faces) . ' faces found for clustering'); |
|
| 155 | |||
| 156 | // Cluster is associative array where key is person ID. |
||
| 157 | // Value is array of face IDs. For old clusters, person IDs are some existing person IDs, |
||
| 158 | // and for new clusters is whatever chinese whispers decides to identify them. |
||
| 159 | // |
||
| 160 | 1 | $currentClusters = $this->getCurrentClusters($faces); |
|
| 161 | 1 | $newClusters = $this->getNewClusters($faces); |
|
| 162 | 1 | $this->logInfo(count($newClusters) . ' persons found after clustering'); |
|
| 163 | |||
| 164 | // New merge |
||
| 165 | 1 | $mergedClusters = $this->mergeClusters($currentClusters, $newClusters); |
|
| 166 | 1 | $this->personMapper->mergeClusterToDatabase($userId, $currentClusters, $mergedClusters); |
|
| 167 | |||
| 168 | // Remove all orphaned persons (those without any faces) |
||
| 169 | // NOTE: we will do this for all models, not just for current one, but this is not problem. |
||
| 170 | 1 | $orphansDeleted = $this->personMapper->deleteOrphaned($userId); |
|
| 171 | 1 | if ($orphansDeleted > 0) { |
|
| 172 | $this->logInfo('Deleted ' . $orphansDeleted . ' persons without faces'); |
||
| 173 | } |
||
| 174 | |||
| 175 | // Prevents not create/recreate the clusters unnecessarily. |
||
| 176 | |||
| 177 | 1 | $this->settingsService->setNeedRecreateClusters(false, $userId); |
|
| 178 | 1 | $this->settingsService->_setForceCreateClusters(false, $userId); |
|
| 179 | 1 | } |
|
| 180 | |||
| 181 | /** |
||
| 182 | * Evaluate whether we want to recreate clusters. We want to recreate clusters/persons if: |
||
| 183 | * - Some cluster/person is invalidated (is_valid is false for someone) |
||
| 184 | * - This means some image that belonged to this user is changed, deleted etc. |
||
| 185 | * - There are some new faces. Now, we don't want to jump the gun here. We want to either have: |
||
| 186 | * - more than 25 new faces, or |
||
| 187 | * - less than 25 new faces, but they are older than 2h |
||
| 188 | * |
||
| 189 | * (basically, we want to avoid recreating cluster for each new face being uploaded, |
||
| 190 | * however, we don't want to wait too much as clusters could be changed a lot) |
||
| 191 | */ |
||
| 192 | private function hasNewFacesToRecreate($userId, $modelId): bool { |
||
| 217 | } |
||
| 218 | |||
| 219 | private function hasStalePersonsToRecreate($userId, $modelId): bool { |
||
| 220 | return $this->personMapper->countPersons($userId, $modelId, true) > 0; |
||
| 221 | } |
||
| 222 | |||
| 223 | private function needRecreateBySettings($userId): bool { |
||
| 224 | return $this->settingsService->getNeedRecreateClusters($userId); |
||
| 225 | } |
||
| 226 | |||
| 227 | 1 | private function needCreateFirstTime($userId, $modelId): bool { |
|
| 228 | // User should not be able to use this directly, used in tests |
||
| 229 | 1 | if ($this->settingsService->_getForceCreateClusters($userId)) |
|
| 230 | 1 | return true; |
|
| 231 | |||
| 232 | 1 | $imageCount = $this->imageMapper->countUserImages($userId, $modelId); |
|
| 233 | 1 | if ($imageCount === 0) |
|
| 234 | return false; |
||
| 235 | |||
| 236 | 1 | $imageProcessed = $this->imageMapper->countUserImages($userId, $modelId, true); |
|
| 237 | 1 | if ($imageProcessed === 0) |
|
| 238 | 1 | return false; |
|
| 239 | |||
| 240 | // These are basic criteria without which we should not even consider creating clusters. |
||
| 241 | // These clusters will be small and not "stable" enough and we should better wait for more images to come. |
||
| 242 | // todo: get rid of magic numbers (move to config) |
||
| 243 | $facesCount = $this->faceMapper->countFaces($userId, $modelId); |
||
| 244 | if ($facesCount > 1000) |
||
| 245 | return true; |
||
| 246 | |||
| 247 | $percentImagesProcessed = $imageProcessed / floatval($imageCount); |
||
| 248 | if ($percentImagesProcessed > 0.95) |
||
| 249 | return true; |
||
| 250 | |||
| 251 | return false; |
||
| 252 | } |
||
| 253 | |||
| 254 | 1 | private function getCurrentClusters(array $faces): array { |
|
| 265 | } |
||
| 266 | |||
| 267 | 1 | private function getNewClusters(array $faces): array { |
|
| 268 | // Clustering parameters |
||
| 269 | 1 | $sensitivity = $this->settingsService->getSensitivity(); |
|
| 270 | 1 | $min_confidence = $this->settingsService->getMinimumConfidence(); |
|
| 271 | 1 | $min_face_size = $this->settingsService->getMinimumFaceSize(); |
|
| 272 | |||
| 273 | // Create edges for chinese whispers |
||
| 274 | 1 | $edges = array(); |
|
| 275 | |||
| 276 | 1 | if (version_compare(phpversion('pdlib'), '1.0.2', '>=')) { |
|
| 277 | $faces_count = count($faces); |
||
| 278 | for ($i = 0; $i < $faces_count; $i++) { |
||
| 279 | $face1 = $faces[$i]; |
||
| 280 | if (($face1->confidence < $min_confidence) || |
||
| 281 | (max($face1->height(), $face1->width()) < $min_face_size)) { |
||
| 282 | $edges[] = array($i, $i); |
||
| 283 | continue; |
||
| 284 | } |
||
| 285 | for ($j = $i; $j < $faces_count; $j++) { |
||
| 286 | $face2 = $faces[$j]; |
||
| 287 | if (($face2->confidence < $min_confidence) || |
||
| 288 | (max($face2->height(), $face2->width()) < $min_face_size)) { |
||
| 289 | continue; |
||
| 290 | } |
||
| 291 | $distance = dlib_vector_length($face1->descriptor, $face2->descriptor); |
||
|
Bug
introduced
by
|
|||
| 292 | if ($distance < $sensitivity) { |
||
| 293 | $edges[] = array($i, $j); |
||
| 294 | } |
||
| 295 | } |
||
| 296 | } |
||
| 297 | } else { |
||
| 298 | 1 | $euclidean = new Euclidean(); |
|
| 299 | 1 | $faces_count = count($faces); |
|
| 300 | 1 | for ($i = 0; $i < $faces_count; $i++) { |
|
| 301 | 1 | $face1 = $faces[$i]; |
|
| 302 | 1 | if (($face1->confidence < $min_confidence) || |
|
| 303 | 1 | (max($face1->height(), $face1->width()) < $min_face_size)) { |
|
| 304 | $edges[] = array($i, $i); |
||
| 305 | continue; |
||
| 306 | } |
||
| 307 | 1 | for ($j = $i; $j < $faces_count; $j++) { |
|
| 308 | 1 | $face2 = $faces[$j]; |
|
| 309 | 1 | if (($face2->confidence < $min_confidence) || |
|
| 310 | 1 | (max($face2->height(), $face2->width()) < $min_face_size)) { |
|
| 311 | continue; |
||
| 312 | } |
||
| 313 | // todo: can't this distance be a method in $face1->distance($face2)? |
||
| 314 | 1 | $distance = $euclidean->distance($face1->descriptor, $face2->descriptor); |
|
| 315 | 1 | if ($distance < $sensitivity) { |
|
| 316 | 1 | $edges[] = array($i, $j); |
|
| 317 | } |
||
| 318 | } |
||
| 319 | } |
||
| 320 | } |
||
| 321 | |||
| 322 | 1 | $newChineseClustersByIndex = dlib_chinese_whispers($edges); |
|
| 323 | 1 | $newClusters = array(); |
|
| 324 | 1 | for ($i = 0, $c = count($newChineseClustersByIndex); $i < $c; $i++) { |
|
| 325 | 1 | if (!isset($newClusters[$newChineseClustersByIndex[$i]])) { |
|
| 326 | 1 | $newClusters[$newChineseClustersByIndex[$i]] = array(); |
|
| 327 | } |
||
| 328 | 1 | $newClusters[$newChineseClustersByIndex[$i]][] = $faces[$i]->id; |
|
| 329 | } |
||
| 330 | |||
| 331 | 1 | return $newClusters; |
|
| 332 | } |
||
| 333 | |||
| 334 | /** |
||
| 335 | * todo: only reason this is public is because of tests. Go figure it out better. |
||
| 336 | */ |
||
| 337 | 3 | public function mergeClusters(array $oldCluster, array $newCluster): array { |
|
| 396 | } |
||
| 397 | } |
||
| 398 |
Loading history...
