arXiv subjects

Anshuman Sinha

Publications and source records attributed to Anshuman Sinha.

10 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

Inverse 3D Microscopy Rendering for Cell Shape Inference with Active Mesh

Traditional methods for biological shape inference, such as deep learning (DL) and active contour models, face important limitations in 3D. DL approaches require large annotated datasets, which are often impractical to obtain, while active contour methods depend on carefully tuned heuristics for intensity attraction and shape regularization. We introduce deltaMic, a novel differentiable 3D renderer for fluorescence microscopy that formulates shape inference as an inverse problem. By leveraging differentiable convolutions, deltaMic simulates the image formation process, integrating a parameterized point spread function (PSF) with a triangle mesh-based representation of biological structures. Unlike DL- or contour-based segmentation, deltaMic directly optimizes both shape and optical parameters to align synthetic and real microscopy images, removing the need for large datasets or sample-specific fine-tuning. To ensure scalability, we implement a GPU-accelerated Fourier transform for triangle meshes along with narrow-band spectral filtering. We show that deltaMic accurately reconstructs cell geometries from both synthetic and diverse experimental 3D microscopy data, while remaining robust to noise and initialization. This establishes a new physics-informed framework for biophysical image analysis and inverse modeling.

physics.bio-ph↗

arXiv2025

Enhancing Audio-Language Models through Self-Supervised Post-Training with Text-Audio Pairs

Research on multi-modal contrastive learning strategies for audio and text has rapidly gained interest. Contrastively trained Audio-Language Models (ALMs), such as CLAP, which establish a unified representation across audio and language modalities, have enhanced the efficacy in various subsequent tasks by providing good text aligned audio encoders and vice versa. These improvements are evident in areas like zero-shot audio classification and audio retrieval, among others. However, the ability of these models to understand natural language and temporal relations is still a largely unexplored and open field for research. In this paper, we propose to equip the multi-modal ALMs with temporal understanding without loosing their inherent prior capabilities of audio-language tasks with a temporal instillation method TeminAL. We implement a two-stage training scheme TeminAL A $\&$ B, where the model first learns to differentiate between multiple sounds in TeminAL A, followed by a phase that instills a sense of time, thereby enhancing its temporal understanding in TeminAL B. This approach results in an average performance gain of $5.28\%$ in temporal understanding on the ESC-50 dataset, while the model remains competitive in zero-shot retrieval and classification tasks on the AudioCap/Clotho datasets. We also note the lack of proper evaluation techniques for contrastive ALMs and propose a strategy for evaluating ALMs in zero-shot settings. The general-purpose zero-shot model evaluation strategy ZSTE, is used to evaluate various prior models. ZSTE demonstrates a general strategy to evaluate all ZS contrastive models. The model trained with TeminAL successfully outperforms current models on most downstream tasks.

cs.SD↗

arXiv2024

Representation-space diffusion models for generating periodic materials

Generative models hold the promise of significantly expediting the materials design process when compared to traditional human-guided or rule-based methodologies. However, effectively generating high-quality periodic structures of materials on limited but diverse datasets remains an ongoing challenge. Here we propose a novel approach for periodic structure generation which fully respect the intrinsic symmetries, periodicity, and invariances of the structure space. Namely, we utilize differentiable, physics-based, structural descriptors which can describe periodic systems and satisfy the necessary invariances, in conjunction with a denoising diffusion model which generates new materials within this descriptor or representation space. Reconstruction is then performed on these representations using gradient-based optimization to recover the corresponding Cartesian positions of the crystal structure. This approach differs significantly from current methods by generating materials in the representation space, rather than in the Cartesian space, which is made possible using an efficient reconstruction algorithm. Consequently, known issues with respecting periodic boundaries and translational and rotational invariances during generation can be avoided, and the model training process can be greatly simplified. We show this approach is able to provide competitive performance on established benchmarks compared to current state-of-the-art methods.

cond-mat.mtrl-sci↗

arXiv2024

Neural networks can be FLOP-efficient integrators of 1D oscillatory integrands

We demonstrate that neural networks can be FLOP-efficient integrators of one-dimensional oscillatory integrands. We train a feed-forward neural network to compute integrals of highly oscillatory 1D functions. The training set is a parametric combination of functions with varying characters and oscillatory behavior degrees. Numerical examples show that these networks are FLOP-efficient for sufficiently oscillatory integrands with an average FLOP gain of 1000 FLOPs. The network calculates oscillatory integrals better than traditional quadrature methods under the same computational budget or number of floating point operations. We find that feed-forward networks of 5 hidden layers are satisfactory for a relative accuracy of 0.001. The computational burden of inference of the neural network is relatively small, even compared to inner-product pattern quadrature rules. We postulate that our result follows from learning latent patterns in the oscillatory integrands that are otherwise opaque to traditional numerical integrators.

cs.LG↗

arXiv2023

The Optimal use of Segmentation for Sampling Calorimeters

One of the key design choices of any sampling calorimeter is how fine to make the longitudinal and transverse segmentation. To inform this choice, we study the impact of calorimeter segmentation on energy reconstruction. To ensure that the trends are due entirely to hardware and not to a sub-optimal use of segmentation, we deploy deep neural networks to perform the reconstruction. These networks make use of all available information by representing the calorimeter as a point cloud. To demonstrate our approach, we simulate a detector similar to the forward calorimeter system intended for use in the ePIC detector, which will operate at the upcoming Electron Ion Collider. We find that for the energy estimation of isolated charged pion showers, relatively fine longitudinal segmentation is key to achieving an energy resolution that is better than 10% across the full phase space. These results provide a valuable benchmark for ongoing EIC detector optimizations and may also inform future studies involving high-granularity calorimeters in other experiments at various facilities.

physics.ins-det↗

arXiv2023

Computational study of non-isothermal slag eye formation and its effects on ladle refining

Ladle refining is one of the most important aspects of high-quality steel production. Ladle argon purging which facilitates the refining process also leads to the unwarranted opening of the slag cover known as Slag Eye-opening and has a deleterious effect on the quality of steel. Slag eye-opening has been analysed in past under isothermal conditions whereas ladle refining is a transient and non-isothermal operation. The current study deals with the modelling of slag-eye opening and its effects on ladle refining under non-isothermal conditions. The bubble plume is modelled with the help of Discrete Phase modelling (DPM) coupled with a discrete random walk model for including the particle level turbulence. Temperature-dependent thermophysical properties of slag are obtained from FactSage. Opening of slag-metal interface cools the slag-eye region, which causes changes in the thermophysical properties of the slag phase. These changes are then reflected in the flow characteristics of this complex fluid. The slags flow profile and eye formation are compared and explained between cold modelling techniques and actual ladle metallurgy. The consequences of changing thermophysical prop during ladle refining manifest in their influence on the overall mass transfer coefficient and the kinetics of desulfurization. This can be achieved without the requirement to solve computationally demanding species transport equations, thereby enhancing the practical efficiency of this approach.

physics.flu-dyn↗

arXiv2010

A Survey of System Security in Contactless Electronic Passports

A traditional paper-based passport contains a Machine- Readable Zone (MRZ) and a Visual Inspection Zone (VIZ). The MRZ has two lines of the holder's personal data, some document data, and verification characters encoded using the Optical Character Recognition font B (OCRB). The encoded data includes the holder's name, date of birth, and other identifying information for the holder or the document. The VIZ contains the holder's photo and signature, usually on the data page. However, the MRZ and VIZ can be easily duplicated with normal document reproduction technology to produce a fake passport which can pass traditional verification. Neither of these features actively verify the holder's identity; nor do they bind the holder's identity to the document. A passport also contains pages for stamps of visas and of country entry and exit dates, which can be easily altered to produce fake permissions and travel records. The electronic passport, supporting authentication using secure credentials on a tamper-resistant chip, is an attempt to improve on the security of the paper-based passport at minimum cost. This paper surveys the security mechanisms built into the firstgeneration of authentication mechanisms and compares them with second-generation passports. It analyzes and describes the cryptographic protocols used in Basic Access Control (BAC) and Extended Access Control (EAC).

cs.CR↗

arXiv2009

Study of Proposed Methods for Improving TCP Performance Over Wireless Links

TCP is designed for networks with assumption that major losses occur only due to congestion of network traffic. On a wireless network TCP misinterprets the transmission losses due to bit errors and handoffs as losses caused by congestion, and triggers congestion control mechanisms. Because of its end to end delivery model, congestion handling and avoidance mechanisms, TCP has been widely accepted as Transport layer protocol for internetworks. Extension of Internetworks over wireless links is inevitable with the spread of ubiquitous computing and mobile communications. This paper presents study of different mechanisms proposed to extend Transport Control Protocol and other alternate solutions to enhance end to end performance over lossy wireless links. The paper studies details of different design choices proposed and their technical advantages and disadvantages. Finally, an analysis and proposal for best choice of proposed schemes are made for wireless networks.

cs.NI↗

arXiv2008

Domain Specific Software Architecture for Design Center Automation

Domain specific software architecture aims at software reuse through construction of domain architecture reference model. The constructed reference model presents a set of individual components and their interaction points. When starting on a new large software project, the design engineer starts with pre-constructed model, which can be easily browsed and picks up opportunities of use in the new solution design. This report discusses application of domain reference design methods by deriving domain specific reference architecture for a product ordering system in a design center. The product in this case is instock and special order blinds from different manufacturers in a large supply store. The development of mature domain specific reference software architecture for this domain is not the objective of this report. However, this report would like to capture the method used in one such process and that is the primary concern of this report. This report lists subjective details of such a process applied to the domain of ordering custom and instock blinds from a large home construction and goods supply store. This report also describes the detailed process of derivation of knowledge models, unified knowledge models and the reference architecture for this domain. However, this domain model is only partially complete which may not be used for any real applications. This report is a result of a course project undertaken while studying this methodology.

cs.SE↗