| KEYWORD |
Image labelling for Artificial Intelligence algorithms optimization
keywords ARTIFICIAL INTELLIGENCE, DEEP LEARNING, COMPUTER V, DATA PROCESSING, IMAGE ANALYSIS
Reference persons SANDRO MOOS, LUCA ULRICH, ENRICO VEZZETTI
Research Groups 3D LAB
Thesis type CASE STUDY BASED
Description Accurate image labeling is a crucial step for training neural networks in various application fields, such as object recognition, medical diagnostics, autonomous driving, and much more. This process involves assigning descriptive tags to images to provide structured data that neural networks can be trained on. The quality of the labels directly affects the performance of the neural network, making the adoption of effective and reliable methodologies essential.
This thesis aims to improve the performance of a neural network for the personalized analysis of the physical ergonomics of workers according to the principles of Industry 5.0.
Deadline 11/07/2025
PROPONI LA TUA CANDIDATURA