Skip to main content
Optimizing Multi-Modal Models for Image-Based Shape Retrieval: The Role of Pre-Alignment and Hard Contrastive Learning | Signal Canvas | ScienceToStartup