Search

Human pose estimation, a vital task in computer vision, involves detecting and localising human joints in images and videos. While single-frame pose estimation has seen significant progress, it often fails to capture the temporal dynamics for understanding complex, continuous movements. We propose Poseidon, a novel multi-frame pose estimation architecture that extends the ViTPose model by integrating temporal information for enhanced accuracy and robustness to address these limitations. Poseidon introduces key innovations: (1) an Adaptive Frame Weighting (AFW) mechanism that dynamically prioritises frames based on their relevance, ensuring that the model focuses on the most informative data; (2) a Multi-Scale Feature Fusion (MSFF) module that aggregates features from different backbone layers to capture both fine-grained details and high-level semantics; and (3) a Cross-Attention module for effective information exchange between central and contextual frames, enhancing the model's temporal coherence. The proposed architecture improves performance in complex video scenarios and offers scalability and computational efficiency suitable for real-world applications. Our approach achieves state-of-the-art performance on the PoseTrack21 and PoseTrack18 datasets, achieving mAP scores of 88.3 and 87.8, respectively, outperforming existing methods.

In this letter, we revisit the symmetries of Galilean Electrodynamics (GED) in a spacetime of generic dimension $d+1$. We show that these symmetries are infinitely many, and in $d=3$ they correspond to the conformal Milne algebra extended by an additional generator. We discuss their application in the context of non-relativistic AdS$_5$/CFT$_4$ correspondence., Comment: 7 pages. v2: typos fixed

Text-to-image diffusion generative models can generate high quality images at the cost of tedious prompt engineering. Controllability can be improved by introducing layout conditioning, however existing methods lack layout editing ability and fine-grained control over object attributes. The concept of multi-layer generation holds great potential to address these limitations, however generating image instances concurrently to scene composition limits control over fine-grained object attributes, relative positioning in 3D space and scene manipulation abilities. In this work, we propose a novel multi-stage generation paradigm that is designed for fine-grained control, flexibility and interactivity. To ensure control over instance attributes, we devise a novel training paradigm to adapt a diffusion model to generate isolated scene components as RGBA images with transparency information. To build complex images, we employ these pre-generated instances and introduce a multi-layer composite generation process that smoothly assembles components in realistic scenes. Our experiments show that our RGBA diffusion model is capable of generating diverse and high quality instances with precise control over object attributes. Through multi-layer composition, we demonstrate that our approach allows to build and manipulate images from highly complex prompts with fine-grained control over object appearance and location, granting a higher degree of control than competing methods., Comment: NeurIPS 2024

We show that a novel holographic correspondence appears after taking a suitable stringy non-relativistic limit of the AdS$_5$/CFT$_4$ duality. This correspondence relates string theory in String Newton-Cartan AdS$_5\times$S$^5$ with Galilean Electrodynamics with five scalars defined on the Penrose conformal boundary. As a first test, we match the Killing vectors on the string theory side with the on-shell symmetries of the dual field theory., Comment: 7 pages. Concise letter version of our longer paper arXiv:2403.02379. To appear in PRL

As a potential non-invasive biomarker for ischaemic stroke, intracranial arterial calcification (IAC) could be used for stroke risk assessment on CT head scans routinely acquired for other reasons (e.g. trauma, confusion). Artificial intelligence methods can support IAC scoring, but they have not yet been developed for clinical imaging. Large heterogeneous clinical CT datasets are necessary for the training of such methods, but they exhibit expected and unexpected data anomalies. Using CTs from a large clinical trial, the third International Stroke Trial (IST-3), we propose a pipeline that uses as input non-enhanced CT scans to output regions of interest capturing selected large intracranial arteries for IAC scoring. Our method uses co-registration with templates. We focus on quality control, using information presence along the z-axis of the imaging to group and apply similarity measures (structural similarity index measure) to triage assessment of individual image series. Additionally, we propose superimposing thresholded binary masks of the series to inspect large quantities of data in parallel. We identify and exclude unrecoverable samples and registration failures. In total, our pipeline processes 10,659 CT series, rejecting 4,322 (41%) in the entire process, 1,450 (14% of the total) during quality control, and outputting 6,337 series. Our pipeline enables effective and efficient region of interest localisation for targeted IAC segmentation., Comment: Accepted at the 2nd Data Engineering in Medical Imaging workshop @ MICCAI 2024

Using an effective Lagrangian model inspired by Takami et al. 2015 we qualitatively study the early post-merger of a nearly symmetric binary Neutron Star (BNS) merger event with a non-vanishing ambient fraction of dark matter. For this we first mimic the dynamics of two oscillating Neutron Star (NS) masses in the gravitational potential well as they merge. We parametrize the dynamics and ejecta properties in the coalescence event allowing the formation of a surrounding debris disk that may be containing a non-vanishing dark matter fraction. In order to analyze the possible novel dark contribution, we start from a dark-matter free modellization as a benchmark. Using Monte Carlo Markov Chain (MCMC) techniques we approximately recover the gravitational waveforms, restricted to early post-merger time interval from existing simulations in the CoRe database. Later, we explore the impact of an additional dark viscous fluid under a prescribed velocity dependent force in the Lagrangian and obtain the resulting waveforms and some spectral features originating in the first few ms in the BNS post-merger. Finally we discuss our qualitative findings and its range of validity in light of the prospects of detectability in present or future experimental settings., Comment: 13 pages, 5 figures

This study presents an innovative method for Alzheimer's disease diagnosis using 3D MRI designed to enhance the explainability of model decisions. Our approach adopts a soft attention mechanism, enabling 2D CNNs to extract volumetric representations. At the same time, the importance of each slice in decision-making is learned, allowing the generation of a voxel-level attention map to produce an explainable MRI. To test our method and ensure the reproducibility of our results, we chose a standardized collection of MRI data from the Alzheimer's Disease Neuroimaging Initiative (ADNI). On this dataset, our method significantly outperforms state-of-the-art methods in (i) distinguishing AD from cognitive normal (CN) with an accuracy of 0.856 and Matthew's correlation coefficient (MCC) of 0.712, representing improvements of 2.4% and 5.3% respectively over the second-best, and (ii) in the prognostic task of discerning stable from progressive mild cognitive impairment (MCI) with an accuracy of 0.725 and MCC of 0.443, showing improvements of 10.2% and 20.5% respectively over the second-best. We achieved this prognostic result by adopting a double transfer learning strategy, which enhanced sensitivity to morphological changes and facilitated early-stage AD detection. With voxel-level precision, our method identified which specific areas are being paid attention to, identifying these predominant brain regions: the hippocampus, the amygdala, the parahippocampal, and the inferior lateral ventricles. All these areas are clinically associated with AD development. Furthermore, our approach consistently found the same AD-related areas across different cross-validation folds, proving its robustness and precision in highlighting areas that align closely with known pathological markers of the disease., Comment: 21 pages, 9 figures, 9 tables

Alzheimer's disease is one of the most incisive illnesses among the neurodegenerative ones, and it causes a progressive decline in cognitive abilities that, in the worst cases, becomes severe enough to interfere with daily life. Currently, there is no cure, so an early diagnosis is strongly needed to try and slow its progression through medical treatments. Handwriting analysis is considered a potential tool for detecting and understanding certain neurological conditions, including Alzheimer's disease. While handwriting analysis alone cannot provide a definitive diagnosis of Alzheimer's, it may offer some insights and be used for a comprehensive assessment. The Sigma-lognormal model is conceived for movement analysis and can also be applied to handwriting. This model returns a set of lognormal parameters as output, which forms the basis for the computation of novel and significant features. This paper presents a machine learning approach applied to handwriting features extracted through the sigma-lognormal model. The aim is to develop a support system to help doctors in the diagnosis and study of Alzheimer, evaluate the effectiveness of the extracted features and finally study the relation among them.

In this letter we use a perturbative approach to find the spectrum of non-relativistic strings in the String Newton-Cartan (SNC) AdS$_5\times$S$^5$ spacetime. We perturb the bosonic sector of the action around a BMN-like folded string solution in light-cone gauge. We find strong evidence that the theory is described by a combination of massive and massless free fields in an anti-de Sitter background by showing that interaction terms up to six scalars vanish after field redefinitions., Comment: 11 pages

We construct a new type of holographic correspondence between non-relativistic string theory in String Newton-Cartan AdS$_5\times$S$^5$ and Galilean Electrodynamics supplemented with 5 uncharged massless free scalar fields living on the 3+1 dimensional Penrose conformal boundary. In our derivation, we start with Maldacena's setting of a stack of $N$ coincident D3-branes and we show that the near-horizon/decoupling limit commutes with the non-relativistic limit, giving a unique answer both from the gravity and gauge theory perspectives. As a first evidence, we systematically compute the Killing vectors on the string theory side, and find that they match the symmetries of the dual gauge theory., Comment: v2: 32 pages, 1 figure. Symmetries match off-shell. Added section 2.1.1 and references. Version accepted for publication in PRD. v3: typos corrected

Using carefully chosen projections, we consider different Carroll limits of relativistic Dirac fermions in any spacetime dimensions. These limits define Carroll fermions of two types: electric and magnetic. The latter type transforms as a reducible but indecomposable representation of the Carroll group. We also build action principles for all Carroll fermions we introduce; in particular, in even dimensions we provide an action principle for a minimal magnetic Carroll fermion, having the same number of components as a Dirac spinor. We then explore the coupling of these fermions to magnetic Carroll gravity in both its first-order and second-order formulations., Comment: 30 pages. References and minor clarifications added

Computed Tomography (CT) is commonly used to image acute ischemic stroke (AIS) patients, but its interpretation by radiologists is time-consuming and subject to inter-observer variability. Deep learning (DL) techniques can provide automated CT brain scan assessment, but usually require annotated images. Aiming to develop a DL method for AIS using labelled but not annotated CT brain scans from patients with AIS, we designed a convolutional neural network-based DL algorithm using routinely-collected CT brain scans from the Third International Stroke Trial (IST-3), which were not acquired using strict research protocols. The DL model aimed to detect AIS lesions and classify the side of the brain affected. We explored the impact of AIS lesion features, background brain appearances, and timing on DL performance. From 5772 unique CT scans of 2347 AIS patients (median age 82), 54% had visible AIS lesions according to expert labelling. Our best-performing DL method achieved 72% accuracy for lesion presence and side. Lesions that were larger (80% accuracy) or multiple (87% accuracy for two lesions, 100% for three or more), were better detected. Follow-up scans had 76% accuracy, while baseline scans 67% accuracy. Chronic brain conditions reduced accuracy, particularly non-stroke lesions and old stroke lesions (32% and 31% error rates respectively). DL methods can be designed for AIS lesion detection on CT using the vast quantities of routinely-collected CT brain scan data. Ultimately, this should lead to more robust and widely-applicable methods., Comment: 12 pages, 5 figures

Despite the large amount of brain CT data generated in clinical practice, the availability of CT datasets for deep learning (DL) research is currently limited. Furthermore, the data can be insufficiently or improperly prepared for machine learning and thus lead to spurious and irreproducible analyses. This lack of access to comprehensive and diverse datasets poses a significant challenge for the development of DL algorithms. In this work, we propose a complete semi-automatic pipeline to address the challenges of preparing a clinical brain CT dataset for DL analysis and describe the process of standardising this heterogeneous dataset. Challenges include handling image sets with different orientations (axial, sagittal, coronal), different image types (to view soft tissues or bones) and dimensions, and removing redundant background. The final pipeline was able to process 5,868/10,659 (45%) CT image datasets. Reasons for rejection include non-axial data (n=1,920), bone reformats (n=687), separated skull base/vault images (n=1,226), and registration failures (n=465). Further format adjustments, including image cropping, resizing and scaling are also needed for DL processing. Of the axial scans that were not localisers, bone reformats or split brains, 5,868/6,333 (93%) were accepted, while the remaining 465 failed the registration process. Appropriate preparation of medical imaging datasets for DL is a costly and time-intensive process., Comment: 9 pages, 5 figures

Segmentation masks of pathological areas are useful in many medical applications, such as brain tumour and stroke management. Moreover, healthy counterfactuals of diseased images can be used to enhance radiologists' training files and to improve the interpretability of segmentation models. In this work, we present a weakly supervised method to generate a healthy version of a diseased image and then use it to obtain a pixel-wise anomaly map. To do so, we start by considering a saliency map that approximately covers the pathological areas, obtained with ACAT. Then, we propose a technique that allows to perform targeted modifications to these regions, while preserving the rest of the image. In particular, we employ a diffusion model trained on healthy samples and combine Denoising Diffusion Probabilistic Model (DDPM) and Denoising Diffusion Implicit Model (DDIM) at each step of the sampling process. DDPM is used to modify the areas affected by a lesion within the saliency map, while DDIM guarantees reconstruction of the normal anatomy outside of it. The two parts are also fused at each timestep, to guarantee the generation of a sample with a coherent appearance and a seamless transition between edited and unedited parts. We verify that when our method is applied to healthy samples, the input images are reconstructed without significant modifications. We compare our approach with alternative weakly supervised methods on the task of brain lesion segmentation, achieving the highest mean Dice and IoU scores among the models considered., Comment: 13 pages, 7 figures. Accepted to IEEE Transactions on Medical Imaging on 6th September 2024

Using kinematic properties of handwriting to support the diagnosis of neurodegenerative disease is a real challenge: non-invasive detection techniques combined with machine learning approaches promise big steps forward in this research field. In literature, the tasks proposed focused on different cognitive skills to elicitate handwriting movements. In particular, the meaning and phonology of words to copy can compromise writing fluency. In this paper, we investigated how word semantics and phonology affect the handwriting of people affected by Alzheimer's disease. To this aim, we used the data from six handwriting tasks, each requiring copying a word belonging to one of the following categories: regular (have a predictable phoneme-grapheme correspondence, e.g., cat), non-regular (have atypical phoneme-grapheme correspondence, e.g., laugh), and non-word (non-meaningful pronounceable letter strings that conform to phoneme-grapheme conversion rules). We analyzed the data using a machine learning approach by implementing four well-known and widely-used classifiers and feature selection. The experimental results showed that the feature selection allowed us to derive a different set of highly distinctive features for each word type. Furthermore, non-regular words needed, on average, more features but achieved excellent classification performance: the best result was obtained on a non-regular, reaching an accuracy close to 90%.

We find the connection between relativistic and non-relativistic string vacua in AdS$_5 \times$S$^5$ in terms of a free parameter $c$ flow. First, we show that the famous relativistic BMN vacuum flows in the large $c$ parameter to an unphysical solution of the non-relativistic theory. Then, we consider the simplest non-relativistic vacuum, found in arXiv:2109.13240 (called BMN-like), and we identify its relativistic origin, namely a non-compact version of the folded string with zero spin, ignored in the past due to its infinite energy. We show that, once the critical closed B-field required by the non-relativistic limit is included, the total energy of such relativistic solution is finite, and in the large $c$ parameter it precisely matches the one of the BMN-like string. We also analyse the case with spin in the transverse AdS directions., Comment: 24 pages, 1 figure; v2, version accepted for publication in Phys. Rev. D

Twenty-six years ago a small committee report, building on earlier studies, expounded a compelling and poetic vision for the future of astronomy, calling for an infrared-optimized space telescope with an aperture of at least $4m$. With the support of their governments in the US, Europe, and Canada, 20,000 people realized that vision as the $6.5m$ James Webb Space Telescope. A generation of astronomers will celebrate their accomplishments for the life of the mission, potentially as long as 20 years, and beyond. This report and the scientific discoveries that follow are extended thank-you notes to the 20,000 team members. The telescope is working perfectly, with much better image quality than expected. In this and accompanying papers, we give a brief history, describe the observatory, outline its objectives and current observing program, and discuss the inventions and people who made it possible. We cite detailed reports on the design and the measured performance on orbit., Comment: Accepted by PASP for the special issue on The James Webb Space Telescope Overview, 29 pages, 4 figures

In some medical imaging tasks and other settings where only small parts of the image are informative for the classification task, traditional CNNs can sometimes struggle to generalise. Manually annotated Regions of Interest (ROI) are sometimes used to isolate the most informative parts of the image. However, these are expensive to collect and may vary significantly across annotators. To overcome these issues, we propose a framework that employs saliency maps to obtain soft spatial attention masks that modulate the image features at different scales. We refer to our method as Adversarial Counterfactual Attention (ACAT). ACAT increases the baseline classification accuracy of lesions in brain CT scans from 71.39% to 72.55% and of COVID-19 related findings in lung CT scans from 67.71% to 70.84% and exceeds the performance of competing methods. We investigate the best way to generate the saliency maps employed in our architecture and propose a way to obtain them from adversarially generated counterfactual images. They are able to isolate the area of interest in brain and lung CT scans without using any manual annotations. In the task of localising the lesion location out of 6 possible regions, they obtain a score of 65.05% on brain CT scans, improving the score of 61.29% obtained with the best competing method., Comment: International Conference on Machine Learning 2023. 17 pages, 7 figures

Spectral curve methods proved to be powerful techniques in the context of relativistic integrable string theories, since they allow to derive the semiclassical spectrum from the minimal knowledge of a Lax pair and a classical string solution. In this paper we initiate the study of the spectral curve for non-relativistic strings in AdS$_5\times S^5$. First we show that for string solutions whose Lax connection is independent of $\sigma$, the eigenvalues of the monodromy matrix do not have any spectral parameter dependence. We remark that this particular behaviour also appears for relativistic strings in flat space. Second, for some simple non-relativistic string solutions where the path ordered exponential of the Lax connection can be computed, we show that the monodromy matrix is either diagonalisable with quasi-momenta independent of the spectral parameter, or non-diagonalisable. For the latter case, we propose a notion of generalised quasi-momenta, based on maximal abelian subalgebras, which retain a dependence on the spectral parameter., Comment: 20 pages

When Gaussian null coordinates are adapted to a Killing horizon, the near-horizon limit is defined by a coordinate rescaling and then by taking the regulator parameter $\varepsilon$ to be small, as a way of zooming into the horizon hypersurface. In this coordinate setting, it is known that the metric of a non-extremal Killing horizon in the near-horizon limit is divergent, and it has been a common practice to impose extremality in order to set the divergent term to zero. Although the metric is divergent, we show for a class of Killing horizons that the vacuum Einstein's equations can be separated into a divergent and a finite part, leading to a well-defined minimal set of Einstein's equations one needs to solve. We extend the result to Einstein gravity minimally coupled to a massless scalar field. We also discuss the case of Einstein gravity coupled to a Maxwell field, in which case the separability holds if the Maxwell potential has non-vanishing components only in the directions of the horizon spatial cross section., Comment: v2: 14 pages, comment added regarding the validity of the result to a class of near-horizon geometries where higher order corrections to equation (1.4) are set to zero. Section 1.3.1 added on applications of the result. Version accepted for publication in Classical and Quantum Gravity