Publications by Marcella Cornia

Tool learning with foundation models aims to endow AI systems with the ability to invoke external resources — such as APIs, computational utilities, and specialized models — to solve complex tasks beyond the reach of standalone language generation. While recent advances in Large Language Models (LLMs) and Multimodal Large Language Models (MLLMs) have expanded their reasoning and perception capabilities, existing tool-use methods are predominantly limited to text-only inputs and closed-world settings. Consequently, they struggle to interpret multimodal user instructions and cannot generalize to tools unseen during training. In this work, we introduce RaTA-Tool, a novel framework for open-world multimodal tool selection. Rather than learning direct mappings from user queries to fixed tool identifiers, our approach enables an MLLM to convert a multimodal query into a structured task description and subsequently retrieve the most appropriate tool by matching this representation against semantically rich, machine-readable tool descriptions. This retrieval-based formulation naturally supports extensibility to new tools without retraining. To further improve alignment between task descriptions and tool selection, we incorporate a preference-based optimization stage using Direct Preference Optimization (DPO). To support research in this setting, we also introduce the first dataset for open-world multimodal tool use, featuring standardized tool descriptions derived from Hugging Face model cards. Extensive experiments demonstrate that our approach significantly improves tool-selection performance, particularly in open-world, multimodal scenarios.

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.

Object-centric models inspired by DETR have become the dominant paradigm for open-vocabulary video instance segmentation (OV-VIS). While recent efforts have reduced the computational cost of pixel decoding, textual modality fusion, and object decoding to make these architectures more suitable for mobile devices, real-time on-device inference at high frame rates remains an open challenge. In this paper, we introduce SegFS, a dual-stream fast-slow framework that significantly improves efficiency without sacrificing accuracy. On sparse keyframes, an open-vocabulary object-based model predicts instance-level representations. These representations are then projected back into the backbone feature space to condition a lightweight fast network, which efficiently relocalizes and segments the instances in subsequent frames. By shifting instance propagation from object decoding to feature-space conditioning, our approach decouples multimodal semantic understanding from dense mask prediction and enables efficient temporal propagation. The proposed fast branch achieves up to 14x lower latency than the mobile-oriented MOBIUS model, while maintaining competitive segmentation performance on standard OV-VIS benchmarks.

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 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.

Image captioning has attracted significant attention within the Computer Vision and Multimedia research domains, resulting in the development of effective methods for generating natural language descriptions of images. Concurrently, the rise of generative models has facilitated the production of highly realistic and high-quality images, particularly through recent advancements in latent diffusion models. In this paper, we propose to leverage the recent advances in Generative AI and create additional training data that can be effectively used to boost the performance of an image captioning model. Specifically, we combine real images with their synthetic counterparts generated by Stable Diffusion using a Mixup data augmentation technique to create novel training examples. Extensive experiments on the COCO dataset demonstrate the effectiveness of our solution in comparison to different baselines and state-of-the-art methods and validate the benefits of using synthetic data to augment the training stage of an image captioning model and improve the quality of the generated captions. Source code and trained models are publicly available at: https://github.com/aimagelab/synthcap_pp.

Multimodal LLMs (MLLMs) are the natural extension of large language models to handle multimodal inputs, combining text and image data. They have recently garnered attention due to their capability to address complex tasks involving both modalities. However, their effectiveness is limited to the knowledge acquired during training, which restricts their practical utility. In this work, we introduce a novel method to enhance the adaptability of MLLMs by integrating external knowledge sources. Our proposed model, Reflective LLaVA (ReflectiVA), utilizes reflective tokens to dynamically determine the need for external knowledge and predict the relevance of information retrieved from an external database. Tokens are trained following a two-stage two-model training recipe. This ultimately enables the MLLM to manage external knowledge while preserving fluency and performance on tasks where external knowledge is not needed. Through our experiments, we demonstrate the efficacy of ReflectiVA for knowledge-based visual question answering, highlighting its superior performance compared to existing methods. Source code and trained models are publicly available at https://github.com/aimagelab/ReflectiVA.

Transformer-based language models like BERT have revolutionized Natural Language Processing (NLP) research, but their application to historical languages remains underexplored. This paper investigates the adaptation of BERT-based embedding models for Latin, a language central to the study of the sacred texts of Christianity. Focusing on Jerome’s Vulgate, pre-Vulgate Latin translations of the Bible, and patristic commentaries such as Augustine’s De Genesi ad litteram, we address the challenges posed by Latin’s complex syntax, specialized vocabulary, and historical variations at the orthographic, morphological, and semantic levels. In particular, we propose fine-tuning existing BERT-based embedding models on annotated Latin corpora, using self-generated hard negatives to improve performance in detecting biblical references in early Christian literature in Latin. Experimental results demonstrate the ability of BERT-based models to identify citations of and allusions to the Bible(s) in ancient Christian commentaries while highlighting the complexities and challenges of this field. By integrating NLP techniques with humanistic expertise, this work provides a case study on intertextual analysis in Latin patristic works. It underscores the transformative potential of interdisciplinary approaches, advancing computational tools for sacred text studies and bridging the gap between philology and computational analysis.

In recent years, the fashion industry has increasingly adopted AI technologies to enhance customer experience, driven by the proliferation of e-commerce platforms and virtual applications. Among the various tasks, virtual try-on and multimodal fashion image editing – which utilizes diverse input modalities such as text, garment sketches, and body poses – have become a key area of research. Diffusion models have emerged as a leading approach for such generative tasks, offering superior image quality and diversity. However, most existing virtual try-on methods rely on having a specific garment input, which is often impractical in real-world scenarios where users may only provide textual specifications. To address this limitation, in this work we introduce Fashion Retrieval-Augmented Generation (Fashion-RAG), a novel method that enables the customization of fashion items based on user preferences provided in textual form. Our approach retrieves multiple garments that match the input specifications and generates a personalized image by incorporating attributes from the retrieved items. To achieve this, we employ textual inversion techniques, where retrieved garment images are projected into the textual embedding space of the Stable Diffusion text encoder, allowing seamless integration of retrieved elements into the generative process. Experimental results on the Dress Code dataset demonstrate that Fashion-RAG outperforms existing methods both qualitatively and quantitatively, effectively capturing fine-grained visual details from retrieved garments. To the best of our knowledge, this is the first work to introduce a retrieval-augmented generation approach specifically tailored for multimodal fashion image editing.