Search

This study addresses the challenge of urban safety in New York City by examining the relationship between the built environment and crime rates using machine learning and a comprehensive dataset of street view images. We aim to identify how urban landscapes correlate with crime statistics, focusing on the characteristics of street views and their association with crime rates. The findings offer insights for urban planning and crime prevention, highlighting the potential of environmental design in enhancing public safety.

Road pavement detection and segmentation are critical for developing autonomous road repair systems. However, developing an instance segmentation method that simultaneously performs multi-class defect detection and segmentation is challenging due to the textural simplicity of road pavement image, the diversity of defect geometries, and the morphological ambiguity between classes. We propose a novel end-to-end method for multi-class road defect detection and segmentation. The proposed method comprises multiple spatial and channel-wise attention blocks available to learn global representations across spatial and channel-wise dimensions. Through these attention blocks, more globally generalised representations of morphological information (spatial characteristics) of road defects and colour and depth information of images can be learned. To demonstrate the effectiveness of our framework, we conducted various ablation studies and comparisons with prior methods on a newly collected dataset annotated with nine road defect classes. The experiments show that our proposed method outperforms existing state-of-the-art methods for multi-class road defect detection and segmentation methods., Comment: Accepted to the ICRA 2024

The advancement of deep learning has facilitated the integration of Artificial Intelligence (AI) into clinical practices, particularly in computer-aided diagnosis. Given the pivotal role of medical images in various diagnostic procedures, it becomes imperative to ensure the responsible and secure utilization of AI techniques. However, the unauthorized utilization of AI for image analysis raises significant concerns regarding patient privacy and potential infringement on the proprietary rights of data custodians. Consequently, the development of pragmatic and cost-effective strategies that safeguard patient privacy and uphold medical image copyrights emerges as a critical necessity. In direct response to this pressing demand, we present a pioneering solution named Medical Image Adversarial watermarking (MIAD-MARK). Our approach introduces watermarks that strategically mislead unauthorized AI diagnostic models, inducing erroneous predictions without compromising the integrity of the visual content. Importantly, our method integrates an authorization protocol tailored for legitimate users, enabling the removal of the MIAD-MARK through encryption-generated keys. Through extensive experiments, we validate the efficacy of MIAD-MARK across three prominent medical image datasets. The empirical outcomes demonstrate the substantial impact of our approach, notably reducing the accuracy of standard AI diagnostic models to a mere 8.57% under white box conditions and 45.83% in the more challenging black box scenario. Additionally, our solution effectively mitigates unauthorized exploitation of medical images even in the presence of sophisticated watermark removal networks. Notably, those AI diagnosis networks exhibit a meager average accuracy of 38.59% when applied to images protected by MIAD-MARK, underscoring the robustness of our safeguarding mechanism.

A p-adic analogue of the pseudonorm version of the birational Torelli type theorem is obtained via a comparison theorem of image closures. Among other results obtained, we have a criterion for existence of rational points of canonically polarized surfaces over finite fields.

We propose a novel hematoxylin and eosin (H&E) stain normalization method based on a modified U-Net neural network architecture. Unlike previous deep-learning methods that were often based on generative adversarial networks (GANs), we take a teacher-student approach and use paired datasets generated by a trained CycleGAN to train a U-Net to perform the stain normalization task. Through experiments, we compared our method to two recent competing methods, CycleGAN and StainNet, a lightweight approach also based on the teacher-student model. We found that our method is faster and can process larger images with better quality compared to CycleGAN. We also compared to StainNet and found that our method delivered quantitatively and qualitatively better results., Comment: Accepted to The 22nd IEEE International Conference on BioInformatics and BioEngineering (BIBE), 2022

In this paper we study the problem of extension of holomorphic sections of line bundles/vector bundles from reduced unions of strata of divisors. An extension theorem of Ohsawa--Takegoshi type is proved. As consequences we deduce several qualitative results on extension from snc divisors and generic global generation of vector bundles., Comment: 28 pages. This version consists of (a revision of) the part about adjoint bundles of the original version; the rest parts about pluricanonical bundles are removed and will appear in another article

We consider the Toda systems of VHS type with singular sources and provide a criterion for the existence of solutions with prescribed asymptotic behaviour near singularities. We also prove the uniqueness of solution. Our approach uses Simpson's theory of constructing Higgs-Hermitian-Yang-Mills metrics from stability.

Given a projective variety X, a smooth divisor D, and semipositive line bundles (L_1,h_1),,...,(L_m,h_m), we consider the "multiply twisted pluricanonical bundle" F:=m(K_X+D)+L_1+...+L_m on X and F_D:=mK_D+(L_1+...+L_m)|_D. Let I_j be the multiplier ideal sheaves associated to h_j, j=1,...,m. We show that, under a certain conditions on curvature, H^0(D,F_D\otimes I_1I_2...I_m) lies in the image of the restriction map H^0(X,F)->H^0(D,F_D). The format of our result is inspired both by Paun's simplification of Siu's proof of invariance of plurigenera and an earlier similar result due to Demailly. The main ingredient is a modification of Siu-Paun's induction construction and an extension theorem of Ohsawa-Takegoshi type (O-T). We also include a detail proof of O-T. The key feature is that the ideal sheaf we use is the product of the multiplier ideals associated to the singular metrics h_1,...,h_m, which contains the multiplier ideal sheaf of the product of the metrics h_1\otimes...\otimes h_m., Comment: 26 pages

A classical set of birational invariants of a variety are its spaces of pluricanonical forms and some of their canonically defined subspaces. Each of these vector spaces admits a typical metric structure which is also birationally invariant. These vector spaces so metrized will be referred to as the pseudonormed spaces of the original varieties. A fundamental question is the following: given two mildly singular projective varieties with some of the first variety's pseudonormed spaces being isometric to the corresponding ones of the second variety's, can one construct a birational map between them which induces these isometries? In this work a positive answer to this question is given for varieties of general type. This can be thought of as a theorem of Torelli type for birational equivalence., Comment: 13 pages, to appear in PNAS

Satellite observations show widespread increasing trends of leaf area index (LAI), known as the Earth greening. However, the biophysical impacts of this greening on land surface temperature (LST) remain unclear. Here, we quantify the biophysical impacts of Earth greening on LST from 2000 to 2014 and disentangle the contributions of different factors using a physically based attribution model. We find that 93% of the global vegetated area shows negative sensitivity of LST to LAI increase at the annual scale, especially for semiarid woody vegetation. Further considering the LAI trends (P ≤ 0.1), 30% of the global vegetated area is cooled by these trends and 5% is warmed. Aerodynamic resistance is the dominant factor in controlling Earth greening’s biophysical impacts: The increase in LAI produces a decrease in aerodynamic resistance, thereby favoring increased turbulent heat transfer between the land and the atmosphere, especially latent heat flux.

Amazon forests experienced two severe droughts at the beginning of the 21st century: one in 2005 and the other in 2010. How Amazon forests responded to these droughts is critical for the future of the Earth's climate system. It is only possible to assess Amazon forests' response to the droughts in large areal extent through satellite remote sensing. Here, we used the Multi-Angle Implementation of Atmospheric Correction (MAIAC) Moderate Resolution Imaging Spectroradiometer (MODIS) vegetation index (VI) data to assess Amazon forests' response to droughts, and compared the results with those from the standard (Collection 5 and Collection 6) MODIS VI data. Overall, the MAIAC data reveal more realistic Amazon forests inter-annual greenness dynamics than the standard MODIS data. Our results from the MAIAC data suggest that: (1) the droughts decreased the greenness (i.e., photosynthetic activity) of Amazon forests; (2) the Amazon wet season precipitation reduction induced by El Niño events could also lead to reduced photosynthetic activity of Amazon forests; and (3) in the subsequent year after the water stresses, the greenness of Amazon forests recovered from the preceding decreases. However, as previous research shows droughts cause Amazon forests to reduce investment in tissue maintenance and defense, it is not clear whether the photosynthesis of Amazon forests will continue to recover after future water stresses, because of the accumulated damages caused by the droughts.

The article focuses on the significance of real estate investment trusts (REITs) to investors as they provide them the chance to enhance their investment performance. It discusses the use of mean-variance spanning tests and assess if structural changes could affect the efficiency of adding REITs to benchmark portfolios. The economic importance of REIT investment is realized through an asset allocation.