activity
20242026
collaborators

7 papers

cs.CV2026

Understanding How MLLMs Describe Artworks Using Token Activation Maps

Nicola Fanelli, Pasquale De Marinis, Raffaele Scaringi +3

Multimodal Large Language Models (MLLMs) describe artworks with remarkable fluency, yet the visual reasoning behind their outputs remains opaque. When an MLLM names a style, identi…

cs.CV2026

Take a Peek: Efficient Encoder Adaptation for Few-Shot Semantic Segmentation via LoRA

Pasquale De Marinis, Gennaro Vessio, Giovanna Castellano

Few-shot semantic segmentation (FSS) aims to segment novel classes in query images using only a small annotated support set. While prior research has mainly focused on improving de…

cs.AI2026

Explainable Fuzzy GNNs for Leak Detection in Water Distribution Networks

Qusai Khaled, Pasquale De Marinis, Moez Louati +3

Timely leak detection in water distribution networks is critical for conserving resources and maintaining operational efficiency. Although Graph Neural Networks (GNNs) excel at cap…

cs.CV2025

DistillFSS: Synthesizing Few-Shot Knowledge into a Lightweight Segmentation Model

Pasquale De Marinis, Pieter M. Blok, Uzay Kaymak +3

Cross-Domain Few-Shot Semantic Segmentation (CD-FSS) seeks to segment unknown classes in unseen domains using only a few annotated examples. This setting is inherently challenging:…

cs.CV2025

Matching-Based Few-Shot Semantic Segmentation Models Are Interpretable by Design

Pasquale De Marinis, Uzay Kaymak, Rogier Brussee +2

Few-Shot Semantic Segmentation (FSS) models achieve strong performance in segmenting novel classes with minimal labeled examples, yet their decision-making processes remain largely…

cs.CV2025

Label Anything: Multi-Class Few-Shot Semantic Segmentation with Visual Prompts

Pasquale De Marinis, Nicola Fanelli, Raffaele Scaringi +4

Few-shot semantic segmentation aims to segment objects from previously unseen classes using only a limited number of labeled examples. In this paper, we introduce Label Anything, a…