arXiv subjects

Alessandro Pasqui

Publications and source records attributed to Alessandro Pasqui.

3 recordsLinked to original sources

arXiv2026

VertAX: a differentiable vertex model for learning epithelial tissue mechanics

Epithelial tissues dynamically reshape through local mechanical interactions among cells, a process well captured by vertex models. Yet their many tunable parameters make inference and optimization challenging, motivating computational frameworks that flexibly model and learn tissue mechanics. We introduce VertAX, a differentiable JAX-based framework for vertex-modeling of confluent epithelia. VertAX provides automatic differentiation, GPU acceleration, and end-to-end bilevel optimization for forward simulation, parameter inference, and inverse mechanical design. Users can define arbitrary energy and cost functions in pure Python, enabling seamless integration with machine-learning pipelines. We demonstrate VertAX on three representative tasks: (i) forward modeling of tissue morphogenesis, (ii) mechanical parameter inference, and (iii) inverse design of tissue-scale behaviors. We benchmark three differentiation strategies-automatic differentiation, implicit differentiation, and equilibrium propagation-showing that the latter can approximate gradients using repeated forward, adjoint-free simulations alone, offering a simple route for extending inverse biophysical problems to non-differentiable simulators with limited additional engineering effort.

cs.LG↗

arXiv2025

Self-Supervised Z-Slice Augmentation for 3D Bio-Imaging via Knowledge Distillation

Three-dimensional biological microscopy has significantly advanced our understanding of complex biological structures. However, limitations due to microscopy techniques, sample properties or phototoxicity often result in poor z-resolution, hindering accurate cellular measurements. Here, we introduce ZAugNet, a fast, accurate, and self-supervised deep learning method for enhancing z-resolution in biological images. By performing nonlinear interpolation between consecutive slices, ZAugNet effectively doubles resolution with each iteration. Compared on several microscopy modalities and biological objects, it outperforms competing methods on most metrics. Our method leverages a generative adversarial network (GAN) architecture combined with knowledge distillation to maximize prediction speed without compromising accuracy. We also developed ZAugNet+, an extended version enabling continuous interpolation at arbitrary distances, making it particularly useful for datasets with nonuniform slice spacing. Both ZAugNet and ZAugNet+ provide high-performance, scalable z-slice augmentation solutions for large-scale 3D imaging. They are available as open-source frameworks in PyTorch, with an intuitive Colab notebook interface for easy access by the scientific community.

cs.CV↗

arXiv2002

Constraints on the interaction and self-interaction of dark energy from cosmic microwave background

It is well-known that even high quality cosmic microwave background (CMB) observations are not sufficient on their own to determine the equation of state of the dark energy, due to the effect of the so-called geometric degeneracy at large multipoles and the cosmic variance at small ones. In contrast, we find that CMB data can put tight constraints on another fundamental property of the dark energy, namely its coupling to dark matter. We compare the current high-resolution CMB data to models of dark energy characterized by an inverse power law or exponential potential and by the coupling to dark matter. We determine the curve of degeneracy between the dark energy equation of state and the dimensionless Hubble parameter h and show that even an independent perfect determination of h may be insufficient to distinguish dark energy from a pure cosmological constant with the current dataset. On the other hand, we find that the interaction with dark matter is firmly bounded, regardless of the potential. In terms of the dimensionless ratio βof the dark energy interaction to gravity, we find β<0.16 (95% c.l.). This implies that the effective equation of state between equivalence and tracking has been close to the pure matter equation of state within 1% and that scalar gravity is at least 40 times weaker than tensor gravity. Further, we show that an experiment limited by cosmic variance only, like the Planck mission, can put an upper bound β< 0.05 (95% c.l.). This shows that CMB observations have a strong potentiality not only as a test of cosmic kinematics but also as a gravitational probe.

astro-ph↗