ConsInstancy: learning instance representations for semi-supervised panoptic segmentation of concrete aggregate particles

Abstract

We present a semi-supervised method for panoptic segmentation based on ConsInstancy regularisation, a novel strategy for semi-supervised learning. It leverages completely unlabelled data by enforcing consistency between predicted instance representations and semantic segmentations during training in order to improve the segmentation performance. To this end, we also propose new types of instance representations that can be predicted by one simple forward path through a fully convolutional network (FCN), delivering a convenient and simple-to-train framework for panoptic segmentation. More specifically, we propose the prediction of a three-dimensional instance orientation map as intermediate representation and two complementary distance transform maps as final representation, providing unique instance representations for a panoptic segmentation. We test our method on two challenging data sets of both, hardened and fresh concrete, the latter being proposed by the authors in this paper demonstrating the effectiveness of our approach, outperforming the results achieved by state-of-the-art methods for semi-supervised segmentation. In particular, we are able to show that by leveraging completely unlabelled data in our semi-supervised approach the achieved overall accuracy (OA) is increased by up to 5% compared to an entirely supervised training using only labelled data. Furthermore, we exceed the OA achieved by state-of-the-art semi-supervised methods by up to 1.5%.

Bildbasierte Frischbetonprüfung – Teil 1: Konsistenz und Leimgehalt des Frischbetons

“…Gemessene Strecken und Flächen in den korrigierten Bildern lassen sich so über einen Skalierungsfaktor in entsprechende Größen im Objektraum umrechnen. Ausführlich beschrieben sind die verwendeten Ansätze und Algorithmen in [21, 22]. Darüber hinaus können weitere Klassen bzw.…”

Bildbasierte Frischbetonprüfung – Teil 2: Granulometrische Eigenschaften der Gesteinskörnung

“…der Gesteinskörnung treffen zu können. Die angewendeten Ansätze und Algorithmen sind ausführlich in [17, 18] beschrieben. Bild 2 zeigt beispielhaft die perspektivisch korrigierte Bildaufnahme eines ausgebreiteten Frischbetons (links) sowie das dazugehörige Binärbild der semantischen Instanz‐Segmentierung der groben Gesteinskörnung (Mitte).…”

Concrete 4.0 ‐ Sustainable concrete construction with digital quality control

“…Finally, a Vision Transformer based method, which is able to model temporal relations within sequential data, is learned as mapping function between the image data and the rheological properties, such as yield stress and plastic viscosity. For more details on the methodology, we refer the reader to [37].…”

“…2). A detailed overview on the developed methodology as well as on the precision of image based methods compared to standard sieve testing is given in [24,25]. Here it could be shown, that using image based methods, the grading curve of aggregates on a conveyor belt as shown in Fig.…”