Publications

Ensuring compliance of face images with ISO/ICAO quality standards is essential for boosting the document enrollment process. Indeed, traditional manual checks are slow, subjective, and difficult to scale. Therefore, we propose a system that aims to fully automate compliance verification by directly analyzing the official requirements without relying on predefined hand-crafted features or manual thresholds. Our method combines a Large Language Model, a novel prompt learning procedure, and a contrastive learning framework to evaluate the adherence of a face image to quality requirements. Tested on a recent dataset, our proposed system achieves high accuracy, surpassing existing academic and commercial solutions. By streamlining the implementation and updates to the compliance rules, our approach represents a significant step toward simple, scalable, and regulation-driven image verification. Code and models are publicly available 1

Attention guides our gaze to fixate the proper location of the scene and holds it in that location for the deserved amount of time given current processing demands, before shifting to the next one. As such, gaze deployment crucially is a temporal process. Existing computational models have made significant strides in predicting spatial aspects of observer's visual scanpaths (where to look), while often putting on the background the temporal facet of attention dynamics (when). In this paper we present TPP-Gaze, a novel and principled approach to model scanpath dynamics based on Neural Temporal Point Process (TPP), that jointly learns the temporal dynamics of fixations position and duration, integrating deep learning methodologies with point process theory. We conduct extensive experiments across five publicly available datasets. Our results show the overall superior performance of the proposed model compared to state-of-the-art approaches.

Multimodal Large Language Models (MLLMs) have recently emerged as a powerful framework for extending the capabilities of Large Language Models (LLMs) to reason over non-textual modalities. However, despite their success, understanding how they integrate visual and textual information remains an open challenge. Among them, LLaMA~3.2-Vision represents a significant milestone in the development of open-source MLLMs, offering a reproducible and efficient architecture that competes with leading proprietary models, such as Claude 3 Haiku and GPT-4o mini. Motivated by these characteristics, we conduct the first systematic analysis of the information flow between vision and language in LLaMA~3.2-Vision. We analyze three visual question answering (VQA) benchmarks, covering the tasks of VQA on natural images---using both open-ended and multiple-choice question formats---as well as document VQA. These tasks require diverse reasoning capabilities, making them well-suited to reveal distinct patterns in multimodal reasoning. Our analysis unveils a four-stage reasoning strategy: an initial semantic interpretation of the question, an early-to-mid-layer multimodal fusion, a task-specific reasoning stage guided by the resulting multimodal embedding, and a final answer prediction stage. Furthermore, we reveal that multimodal fusion is task-dependent: in complex settings such as document VQA, the model postpones cross-modal integration until semantic reasoning over the question has been established. Overall, our findings offer new insights into the internal dynamics of MLLMs and contribute to advancing the interpretability of vision-language architectures. Our source code is available at https://github.com/AImageLab/MLLMs-FlowTracker.

Trajectory forecasting is crucial for video surveillance analytics, as it enables the anticipation of future movements for a set of agents, e.g., basketball players engaged in intricate interactions with long-term intentions. Deep generative models offer a natural learning approach for trajectory forecasting, yet they encounter difficulties in achieving an optimal balance between sampling fidelity and diversity. We address this challenge by leveraging Vector Quantized Variational Autoencoders (VQ-VAEs), which utilize a discrete latent space to tackle the issue of posterior collapse. Specifically, we introduce an instance-based codebook that allows tailored latent representations for each example. In a nutshell, the rows of the codebook are dynamically adjusted to reflect contextual information (i.e., past motion patterns extracted from the observed trajectories). In this way, the discretization process gains flexibility, leading to improved reconstructions. Notably, instance-level dynamics are injected into the codebook through low-rank updates, which restrict the customization of the codebook to a lower dimension space. The resulting discrete space serves as the basis of the subsequent step, which regards the training of a diffusion-based predictive model. We show that such a two-fold framework, augmented with instance-level discretization, leads to accurate and diverse forecasts, yielding state-of-the-art performance on three established benchmarks.

Despite their remarkable success in medical image segmentation, the life cycle of deep neural networks remains a challenge in clinical applications. These models must be regularly updated to integrate new medical data and customized to meet evolving diagnostic standards, regulatory requirements, commercial needs, and privacy constraints. Model merging offers a promising solution, as it allows working with multiple specialized networks that can be created and combined dynamically instead of relying on monolithic models. While extensively studied in standard 2D classification, the potential of model merging for 3D segmentation remains unexplored. This paper presents an efficient framework that allows effective model merging in the domain of 3D image segmentation. Our approach builds upon theoretical analysis and encourages wide minima during pre-training, which we demonstrate to facilitate subsequent model merging. Using U-Net 3D, we evaluate the method on distinct anatomical structures with the ToothFairy2 and BTCV Abdomen datasets. To support further research, we release the source code and all the model weights in a dedicated repository: https://github.com/LucaLumetti/UNetTransplant

Can the very fabric of how we visually explore the world hold the key to distinguishing individuals with Autism Spectrum Disorder (ASD)? While eye tracking has long promised quantifiable insights into neurodevelopmental conditions, the causal underpinnings of gaze behaviour remain largely uncharted territory. Moving beyond traditional descriptive metrics of gaze, this study employs cutting-edge causal discovery methods to reconstruct the directed networks that govern the flow of attention across natural scenes. Given the well-documented atypical patterns of visual attention in ASD, particularly regarding socially relevant cues, our central hypothesis is that individuals with ASD exhibit distinct causal signatures in their gaze patterns, significantly different from those of typically developing controls. To our knowledge, this is the first study to explore the diagnostic potential of causal modeling of eye movements in uncovering the cognitive phenotypes of ASD and offers a novel window into the neurocognitive alterations characteristic of the disorder.

Foundation models serve as the backbone for numerous specialized models developed through fine-tuning. However, when the underlying pretrained model is updated or retrained (e.g., on larger and more curated datasets), the fine-tuned model becomes obsolete, losing its utility and requiring retraining. This raises the question: is it possible to transfer fine-tuning to a new release of the model? In this work, we investigate how to transfer fine-tuning to a new checkpoint without having to re-train, in a data-free manner. To do so, we draw principles from model re-basin and provide a recipe based on weight permutations to re-base the modifications made to the original base model, often called task vector. In particular, our approach tailors model re-basin for Transformer models, taking into account the challenges of residual connections and multi-head attention layers. Specifically, we propose a two-level method rooted in spectral theory, initially permuting the attention heads and subsequently adjusting parameters within select pairs of heads. Through extensive experiments on visual and textual tasks, we achieve the seamless transfer of fine-tuned knowledge to new pre-trained backbones without relying on a single training step or datapoint.

Styled Handwritten Text Generation (HTG) has received significant attention in recent years,propelled by the success of learning-based solutions employing GANs,Transformers,and,preliminarily,Diffusion Models. Despite this surge in interest,there remains a critical yet understudied aspect - the impact of the input,both visual and textual,on the HTG model training and its subsequent influence on performance. This work extends the VATr [1] Styled-HTG approach by addressing the pre-processing and training issues that it faces,which are common to many HTG models. In particular,we propose generally applicable strategies for input preparation and training regularization that allow the model to achieve better performance and generalization capabilities. Moreover,in this work,we go beyond performance optimization and address a significant hurdle in HTG research - the lack of a standardized evaluation protocol. In particular,we propose a standardization of the evaluation protocol for HTG and conduct a comprehensive benchmarking of existing approaches. By doing so,we aim to establish a foundation for fair and meaningful comparisons between HTG strategies,fostering progress in the field.

Inference-time scaling has recently gained attention as an effective strategy for improving the performance of generative models without requiring additional training. Although this paradigm has been successfully applied in text and image generation tasks, its extension to video diffusion models remains relatively underexplored. Indeed, video generation presents unique challenges due to its spatiotemporal complexity, particularly in evaluating intermediate generated samples, a procedure that is required by inference-time scaling algorithms. In this work, we systematically investigate the role of the verifier: the scoring mechanism used to guide sampling. We show that current verifiers, when applied at early diffusion steps, face significant reliability challenges due to noisy samples. We further demonstrate that fine-tuning verifiers on partially denoised samples significantly improves early-stage evaluation and leads to gains in generation quality across multiple inference-time scaling algorithms, including Greedy Search, Beam Search, and a novel Successive Halving baseline.