Publications by Rita Cucchiara

Multimodal Large Language Models (MLLMs) have shown impressive capabilities in jointly understanding text, images, and videos, often evaluated via Visual Question Answering (VQA). However, even state-of-the-art MLLMs struggle with domain-specific or knowledge-intensive queries, where relevant information is underrepresented in pre-training data. Knowledge-based VQA (KB-VQA) addresses this by retrieving external documents to condition answer generation, but current retrieval-augmented approaches suffer from low precision, noisy passages, and limited reasoning. To address this, we propose ReAG, a novel Reasoning-Augmented Multimodal RAG approach that combines coarse- and fine-grained retrieval with a critic model that filters irrelevant passages, ensuring high-quality additional context. The model follows a multi-stage training strategy leveraging reinforcement learning to enhance reasoning over retrieved content, while supervised fine-tuning serves only as a cold start. Extensive experiments on Encyclopedic-VQA and InfoSeek demonstrate that ReAG significantly outperforms prior methods, improving answer accuracy and providing interpretable reasoning grounded in retrieved evidence. Our source code is publicly available at: https://github.com/aimagelab/ReAG.

Large-scale benchmarks such as BraTS have driven progress in brain tumor segmentation, but they provide only masks with limited access to the clinical semantics found in radiology reports. We introduce ReportX, a paired resource of 257 clinical reports aligned to BraTS-GLI-2023 subjects, structured into a rich set of qualitative and quantitative attributes. Qualitative fields are curated by clinicians, while quantitative descriptors are automatically derived via atlas-based localization and geometric computations. We compare our annotation schema to existing report-augmented datasets and show that ReportX provides substantially broader coverage of clinically relevant factors. To exploit this supervision, we encode reports using biomedical language models and incorporate their embeddings as auxiliary semantic guidance for 3D tumor segmentation during training. Experimental results demonstrate that the proposed vision-text alignment improves segmentation performance on standard BraTS metrics, with clinically curated reports providing more consistent improvements than automatically generated or less-structured counterparts. We publicly release the dataset (https://ditto.ing.unimore.it/reportx) and the source code (https://github.com/AImageLab-zip/ReportX).

Flow matching models have recently emerged as an efficient alternative to diffusion, especially for text-guided image generation and editing, offering faster inference through continuous-time dynamics. However, existing flow-based editors predominantly support global or single-instruction edits and struggle with multi-instance scenarios, where multiple parts of a reference input must be edited independently without semantic interference. We identify this limitation as a consequence of globally conditioned velocity fields and joint attention mechanisms, which entangle concurrent edits. To address this issue, we introduce Instance-Disentangled Attention, a mechanism that partitions joint attention operations, enforcing binding between instance-specific textual instructions and spatial regions during velocity field estimation. We evaluate our approach on both natural image editing and a newly introduced benchmark of text-dense infographics with region-level editing instructions. Experimental results demonstrate that our approach promotes edit disentanglement and locality while preserving global output coherence, enabling single-pass, instance-level editing.

In the realm of creative expression, not everyone possesses the gift of effortlessly translating their imaginative visions into flawless sketches. More often than not, the outcome resembles an abstract, perhaps even slightly distorted representation. The art of producing impeccable sketches is not only challenging but also a time-consuming process. Our work is the first of this kind in transforming abstract, sometimes deformed garment sketches into photorealistic catalog images, to empower the everyday individual to become their own fashion designer. We create Sketch2Stitch, a dataset featuring over 65,000 abstract sketch images generated from garments of DressCode and VITONHD, two benchmark datasets in the virtual try-on task. Sketch2Stitch is the first dataset in the literature to provide abstract sketches in the fashion domain. We propose a StyleGAN-based generative framework that bridges freehand sketching with photorealistic garment synthesis. We demonstrate that our framework allows users to sketch rough outlines and optionally provide color hints, producing realistic designs in seconds. Experimental results demonstrate, both quantitatively and qualitatively, that the proposed framework achieves superior performance against various baselines and existing methods on both subsets of our dataset. Our work highlights a pathway toward AI-assisted fashion design tools, democratizing garment ideation for students, independent designers, and casual creators.

This report is the transcript of what was discussed in a convention at the Endocrinology Unit in Modena, Italy, in the form of the aporetic dialogs of ancient Greece. It is the third episode of a series of four discussions on the differences between males and females, with a multidisciplinary approach. In this work, the role of testosterone in gender differences in the aptitude for mathematics is explored. First, the definitions of mathematical abilities were provided together with any gender difference in the distribution of females and males in science, technology, engineering, and mathematics subjects. A clear predominance of males is evident at most science, technology, engineering, and mathematics education levels, especially in advanced academic careers. Then, the discussants were divided into two groups: group 1, which illustrated the thesis that testosterone promotes the development of logical‒mathematical skills, and group 2, which, in contrast, asserted the inconsistency of a direct role of testosterone in improving cognitive abilities and that socio-cultural factors should be considered on the basis of this gender gap. In the end, an expert referee (a female engineer) tried to resolve the aporia: are the two theories equivalent or is one superior?.

This study presents an interdisciplinary methodology for detecting biblical references in Latin patristic literature through an innovative combination of rigorous philological approach and Natural Language Processing (NLP) techniques. Focusing on one of the most influential ancient Christian commentaries on the Bible, Augustine of Hippo’s De Genesi ad litteram, and its relationship with Latin biblical texts (specifically, Jerome’s Vulgate and pre-Vulgate versions), this research introduces a token-based classification system for intertextual references, enriched with semantic annotations and supported by the INCEpTION platform. The first section shows how this numerical classification system accounts for exact matches, lemmatized forms, roots, synonyms, and other forms of semantic parallels (here referred to as “structures”), capturing a wide spectrum of textual similarity. To enhance automatic retrieval of these intertextual connections, we fine-tune BERT-based language models for Latin, incorporating contrastive learning and hard negative mining. In the second section, experimental results show that finetuned models significantly outperform baseline models at various levels of textual similarity. This work highlights the utility of computational models in overcoming the traditional dichotomy between explicit quotations and implicit allusions, embracing multiple intermediate nuances of similarity and offering a scalable approach to the study of intertextuality in ancient writings.

Inference-time scaling has emerged as an effective way to improve generative models at test time by using a verifier to score and select candidate outputs. A common choice is to employ Multimodal Large Language Models (MLLMs) as verifiers, which can improve performance but introduce substantial inference-time cost. Indeed, diffusion pipelines operate in an autoencoder latent space to reduce computation, yet MLLM verifiers still require decoding candidates to pixel space and re-encoding them into the visual embedding space, leading to redundant and costly operations. In this work, we propose Verifier on Hidden States (VHS), a verifier that operates directly on intermediate hidden representations of Diffusion Transformer (DiT) single-step generators. VHS analyzes generator features without decoding to pixel space, thereby reducing the per-candidate verification cost while improving or matching the performance of MLLM-based competitors. We show that, under tiny inference budgets with only a small number of candidates per prompt, VHS enables more efficient inference-time scaling reducing joint generation-and-verification time by 63.3%, compute FLOPs by 51% and VRAM usage by 14.5% with respect to a standard MLLM verifier, achieving a +2.7% improvement on GenEval at the same inference-time budget.

Regesta are catalogs of summaries of other documents and, in some cases, are the only source of information about the content of such full-length documents. For this reason, they are of great interest to scholars in many social and humanities fields. In this work, we focus on Regesta Pontificum Romanum, a large collection of papal registers. Regesta are visually rich documents, where the layout is as important as the text content to convey the contained information through the structure, and are inherently multi-page documents. Among Digital Humanities techniques that can help scholars efficiently exploit regesta and other documental sources in the form of scanned documents, Document Parsing has emerged as a task to process document images and convert them into machine-readable structured representations, usually markup language. However, current models focus on scientific and business documents, and most of them consider only single-paged documents. To overcome this limitation, in this work, we propose μgat, an extension of the recently proposed Document parsing Nougat architecture, which can handle elements spanning over the single page limits. Specifically, we adapt Nougat to process a larger, multi-page context, consisting of the previous and the following page, while parsing the current page. Experimental results, both qualitative and quantitative, demonstrate the effectiveness of our proposed approach also in the case of the challenging Regesta Pontificum Romanorum.