Exploring Bubble Flow with Deep Learning: A Preliminary Study

Non programmé
20m
Amphi 1 (Pôle Commun)

Amphi 1

Pôle Commun

Université Clermont Auvergne Campus des Cézeaux, 63170 Aubière

Description

In a bubble flow study, the bubble size measurement requires identifying every instance in each image frame in the first place. However, their tininess and highly overlapped instances hindered not only human-level manipulation but also any deep learning-based detectors. In addition, their dense presentation in the observation zone made manually measuring phase time-consuming and laborious. Annotation pipelines, adopted foundation models, and human intervention were proposed; those accelerate this process and help to build a real-world bubble detection dataset ready to fine-tune a deep learning model faster. Experimental results yield YOLO as the best balancing detector when coupled with upscaling during training for tiny bubble detection and its application in bubble size distribution.

Auteur

Trong Nghia NGO (Equipe Comsee, Axe ISPR, Institu Pascal, UCA)

Co-auteurs

M. Adib Essid (Axe GePEB, Institut Pascal, UCA) Prof. Christophe Vial (Axe GePEB, Institut Pascal, UCA) Dr Jean-Sébastien Guez (Axe GePEB, Institut Pascal, UCA) Prof. Thierry Chateau (Equipe Comsee, Axe ISPR, Institu Pascal, UCA)

Documents de présentation

Aucun document.