memories/lib/Db/TimelineQueryPeopleFaceReco...

242 lines
8.9 KiB
PHP

<?php
declare(strict_types=1);
namespace OCA\Memories\Db;
use OCP\DB\QueryBuilder\IQueryBuilder;
use OCP\Files\Folder;
use OCP\IDBConnection;
trait TimelineQueryPeopleFaceRecognition
{
protected IDBConnection $connection;
public function transformPeopleFaceRecognitionFilter(IQueryBuilder &$query, string $userId, int $currentModel, string $personStr)
{
// Get title and uid of face user
$personNames = explode('/', $personStr);
if (2 !== \count($personNames)) {
throw new \Exception('Invalid person query');
}
$personUid = $personNames[0];
$personName = $personNames[1];
// Join with images
$query->innerJoin('m', 'facerecog_images', 'fri', $query->expr()->andX(
$query->expr()->eq('fri.file', 'm.fileid'),
$query->expr()->eq('fri.model', $query->createNamedParameter($currentModel)),
));
// Join with faces
$query->innerJoin(
'fri',
'facerecog_faces',
'frf',
$query->expr()->eq('frf.image', 'fri.id')
);
// Join with persons
$nameField = is_numeric($personName) ? 'frp.id' : 'frp.name';
$query->innerJoin('frf', 'facerecog_persons', 'frp', $query->expr()->andX(
$query->expr()->eq('frf.person', 'frp.id'),
$query->expr()->eq('frp.user', $query->createNamedParameter($personUid)),
$query->expr()->eq($nameField, $query->createNamedParameter($personName)),
));
}
public function transformPeopleFaceRecognitionRect(IQueryBuilder &$query, string $userId)
{
// Include detection params in response
$query->addSelect(
'frf.x AS face_x',
'frf.y AS face_y',
'frf.width AS face_width',
'frf.height AS face_height',
'm.w AS image_width',
'm.h AS image_height',
);
}
public function getPeopleFaceRecognition(TimelineRoot &$root, int $currentModel, bool $show_clusters = false, bool $show_singles = false, bool $show_hidden = false)
{
$query = $this->connection->getQueryBuilder();
// SELECT all face clusters
$count = $query->func()->count($query->createFunction('DISTINCT m.fileid'), 'count');
$query->select('frp.id', 'frp.user as user_id', 'frp.name', $count)->from('facerecog_persons', 'frp');
// WHERE there are faces with this cluster
$query->innerJoin('frp', 'facerecog_faces', 'frf', $query->expr()->eq('frp.id', 'frf.person'));
// WHERE faces are from images.
$query->innerJoin('frf', 'facerecog_images', 'fri', $query->expr()->eq('fri.id', 'frf.image'));
// WHERE these items are memories indexed photos
$query->innerJoin('fri', 'memories', 'm', $query->expr()->andX(
$query->expr()->eq('fri.file', 'm.fileid'),
$query->expr()->eq('fri.model', $query->createNamedParameter($currentModel)),
));
// WHERE these photos are in the user's requested folder recursively
$query = $this->joinFilecache($query, $root, true, false);
if ($show_clusters) {
// GROUP by ID of face cluster
$query->groupBy('frp.id');
$query->where($query->expr()->isNull('frp.name'));
} else {
// GROUP by name of face clusters
$query->groupBy('frp.name');
$query->where($query->expr()->isNotNull('frp.name'));
}
// By default hides individual faces when they have no name.
if ($show_clusters && !$show_singles) {
$query->having($query->expr()->gt('count', $query->createNamedParameter(1)));
}
// By default it shows the people who were not hidden
if (!$show_hidden) {
$query->andWhere($query->expr()->eq('frp.is_visible', $query->createNamedParameter(true)));
}
// ORDER by number of faces in cluster
$query->orderBy('count', 'DESC');
$query->addOrderBy('name', 'ASC');
$query->addOrderBy('frp.id'); // tie-breaker
// FETCH all faces
$cursor = $this->executeQueryWithCTEs($query);
$faces = $cursor->fetchAll();
// Post process
foreach ($faces as &$row) {
$row['id'] = $row['name'] ?: (int) $row['id'];
$row['count'] = (int) $row['count'];
}
return $faces;
}
public function getFaceRecognitionPreview(TimelineRoot &$root, $currentModel, $previewId)
{
$query = $this->connection->getQueryBuilder();
// SELECT face detections
$query->select(
'fri.file as file_id', // Get actual file
'frf.x', // Image cropping
'frf.y',
'frf.width',
'frf.height',
'm.w as image_width', // Scoring
'm.h as image_height',
'frf.confidence',
'm.fileid',
'm.datetaken', // Just in case, for postgres
)->from('facerecog_faces', 'frf');
// WHERE faces are from images and current model.
$query->innerJoin('frf', 'facerecog_images', 'fri', $query->expr()->andX(
$query->expr()->eq('fri.id', 'frf.image'),
$query->expr()->eq('fri.model', $query->createNamedParameter($currentModel)),
));
// WHERE these photos are memories indexed
$query->innerJoin('fri', 'memories', 'm', $query->expr()->eq('m.fileid', 'fri.file'));
$query->innerJoin('frf', 'facerecog_persons', 'frp', $query->expr()->eq('frp.id', 'frf.person'));
if (is_numeric($previewId)) {
// WHERE faces are from id persons (a cluster).
$query->where($query->expr()->eq('frp.id', $query->createNamedParameter($previewId)));
} else {
// WHERE faces are from name on persons.
$query->where($query->expr()->eq('frp.name', $query->createNamedParameter($previewId)));
}
// WHERE these photos are in the user's requested folder recursively
$query = $this->joinFilecache($query, $root, true, false);
// LIMIT results
$query->setMaxResults(15);
// Sort by date taken so we get recent photos
$query->orderBy('m.datetaken', 'DESC');
$query->addOrderBy('m.fileid', 'DESC'); // tie-breaker
// FETCH face detections
$cursor = $this->executeQueryWithCTEs($query);
$previews = $cursor->fetchAll();
if (empty($previews)) {
return null;
}
// Score the face detections
foreach ($previews as &$p) {
// Get actual pixel size of face
$iw = min((int) ($p['image_width'] ?: 512), 2048);
$ih = min((int) ($p['image_height'] ?: 512), 2048);
// Get percentage position and size
$p['x'] = (float) $p['x'] / $p['image_width'];
$p['y'] = (float) $p['y'] / $p['image_height'];
$p['width'] = (float) $p['width'] / $p['image_width'];
$p['height'] = (float) $p['height'] / $p['image_height'];
$w = (float) $p['width'];
$h = (float) $p['height'];
// Get center of face
$x = (float) $p['x'] + (float) $p['width'] / 2;
$y = (float) $p['y'] + (float) $p['height'] / 2;
// 3D normal distribution - if the face is closer to the center, it's better
$positionScore = exp(-($x - 0.5) ** 2 * 4) * exp(-($y - 0.5) ** 2 * 4);
// Root size distribution - if the image is bigger, it's better,
// but it doesn't matter beyond a certain point
$imgSizeScore = ($iw * 100) ** (1 / 2) * ($ih * 100) ** (1 / 2);
// Faces occupying too much of the image don't look particularly good
$faceSizeScore = (-$w ** 2 + $w) * (-$h ** 2 + $h);
// Combine scores
$p['score'] = $positionScore * $imgSizeScore * $faceSizeScore * $p['confidence'];
}
// Sort previews by score descending
usort($previews, function ($a, $b) {
return $b['score'] <=> $a['score'];
});
return $previews;
}
/** Convert face fields to object */
private function processFaceRecognitionDetection(&$row, $days = false)
{
if (!isset($row)) {
return;
}
// Differentiate Recognize queries from Face Recognition
if (!isset($row['face_width']) || !isset($row['image_width'])) {
return;
}
if (!$days) {
$row['facerect'] = [
// Get percentage position and size
'w' => (float) $row['face_width'] / $row['image_width'],
'h' => (float) $row['face_height'] / $row['image_height'],
'x' => (float) $row['face_x'] / $row['image_width'],
'y' => (float) $row['face_y'] / $row['image_height'],
];
}
unset($row['face_x'], $row['face_y'], $row['face_w'], $row['face_h'], $row['image_height'], $row['image_width']);
}
}