2 papers
cs.LG2026
Silhouette Loss: Differentiable Global Structure Learning for Deep Representations
Matheus VinÃcius Todescato, Joel LuÃs Carbonera
Learning discriminative representations is a central goal of supervised deep learning. While cross-entropy (CE) remains the dominant objective for classification, it does not expli…
cs.CV2025
No Labels Needed: Zero-Shot Image Classification with Collaborative Self-Learning
Matheus VinÃcius Todescato, Joel LuÃs Carbonera
While deep learning, including Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs), has significantly advanced classification performance, its typical reliance on e…