Based Method
Image-based methods are increasingly used across diverse scientific and engineering fields to analyze visual data, aiming to automate tasks and extract meaningful information. Current research focuses on improving accuracy and efficiency through advancements in model architectures like Vision Transformers (ViTs), convolutional neural networks (CNNs), and novel loss functions, often incorporating techniques from explainable AI to enhance interpretability. These methods are proving impactful in various applications, ranging from medical diagnostics (e.g., developmental disorder assessment) and robotics (e.g., autonomous navigation and object manipulation) to agriculture (e.g., plant health monitoring) and industrial processes (e.g., quality control).
Papers
Optimizing Parking Space Classification: Distilling Ensembles into Lightweight Classifiers
Paulo Luza Alves, André Hochuli, Luiz Eduardo de Oliveira, Paulo Lisboa de Almeida
Comparison of marker-less 2D image-based methods for infant pose estimation
Lennart Jahn, Sarah Flügge, Dajie Zhang, Luise Poustka, Sven Bölte, Florentin Wörgötter, Peter B Marschik, Tomas Kulvicius