
Most modern image denoising networks rely on large depth and parameter counts, making limited use of technical knowledge brought by classical model-based denoisers. This poses a challenge not only for understanding their properties, but also for deploying them on real devices which may present resource constraints and diverse noise profiles. Tackling both issues, we propose a RAW-to-RAW denoising architecture that incorporates the structure of classical self-similarity-based denoisers into a fully learnable neural network. Our design centers on a novel nonlocal block that parallels the established pipeline of neighbor matching, collaborative filtering and aggregation popularized by nonlocal patch-based methods, operating on learned multiscale feature representations. This built-in nonlocality efficiently expands the receptive field, so that a single block per scale with a moderate number of neighbors is sufficient to obtain high-quality results. Training the network on a curated dataset with clean real RAW data and modeled synthetic noise while conditioning it on a noise level map yields a sensor-agnostic denoiser that generalizes effectively to unseen devices. Both quantitative and visual results on benchmarks and in-the-wild photographs position our method as a practical solution for real-world RAW denoising, achieving results competitive with state-of-the-art convolutional and transformer-based denoisers while having significantly fewer parameters and a faster runtime. The code is available at https://github.com/MIA-UIB/nonlocal-matchfilter.
Discrete fuzzy numbers, and in particular those defined over a finite chain have been effectively employed to represent linguistic information within the framework of fuzzy systems. Research on total orderings of such types of fuzzy subsets, and specifically those belonging to the set of discrete fuzzy numbers whose support is a closed subinterval of a finite chain and whose membership values belong to a finite set of the unit interval, has facilitated the development of new methods for constructing logical connectives, based on a bijective function, called pos function, that determines the position of each discrete fuzzy number. For this reason, in this work we revisit the problem by introducing algorithms that exploit the combinatorial structure of total orders to compute the pos function and its inverse exactly. Both the ranking and unranking algorithms run in O(n2mlog n) time, where n+1 is the cardinality of the underlying finite chain and m is the number of allowed membership values. Consequently, the complexity is O(n2log n) when m is fixed, whereas it is linear in m when n is fixed. The results demonstrate that this formulation substantially reduces the computational cost and enables the efficient implementation of algebraic operations—such as aggregation and implication functions—on the set of discrete fuzzy numbers.
In an era of increasing concerns over data privacy, particularly regarding Personally Identifiable Information (PII), robust verification mechanisms are essential. This paper presents an identity verification protocol leveraging Soulbound Tokens (SBTs), public-key cryptography, and zero-knowledge proofs (ZKPs) to enable secure and privacy-preserving identity attribute verification. With the definition of a new kind of token, the zkSBT, the proposed protocol allows users to selectively disclose identity-related attributes without revealing sensitive data. The proposal integrates Merkle tree structures for scalable attribute storage and introducew comparator operators for flexible verification (equal, different, greater, less). To prevent replay attacks, a timestamp mechanism is included in the ZKP circuit. The protocol is implemented using Solidity for smart contracts and Circom for ZKPs. The the protocol is evaluated in terms of performance and security. The results show that the protocol’s execution costs are economically viable. Moreover, it maintains critical security guarantees, ensuring data integrity and preventing unauthorized modifications. The proposed approach facilitates trust in decentralized identity systems, with practical applications in digital voting, online banking, and e-commerce.
Herein, we have reported the synthesis and crystal structure of four multicomponent compounds viz. (fumH2)(4-CNPy)2 (1), (succH2)(4-CNPy)2 (2), (2-Cl-3-OHpy)(4,4-bpy)& sdot;0.5H2O (3) and (4-AmpyH)2(ADS)& sdot;3.75H2O (4), (where, fumH2 = fumaric acid, succH2 = succinic acid, 2-Cl-3-OHpy =2-chloro-3-hydroxypyridine, CNpy = cyanopyridine, bpy = bipyridine, Ampy = Aminopyridine and ADS = anthraquinone-1,5-disulfonate). Compound 1 and 2 are co-crystals whereas compound 3 and 4 are co-crystal hydrate and salt hydrate, respectively. The pKa-based approach has been utilized for the synthesis of co-crystals and the molecular salt reported here. Moreover, crystallographic evidences based on the C-N-C bond angle of the pyridine ring and C-O bond distances of dicarboxylic acid co-formers in the respective compounds have also been analyzed to support molecular salt or co-crystal formation. Crystal structure analysis of the co-crystals 1 & 2 showcase structure guiding R22(7) H-bonding synthon, antiparallel C--O & ctdot;C--O interaction and nitrile-nitrile interactions. Unconventional Type I Cl & ctdot;Cl contact along with hydrogen bonding and pi-stacking interaction strengthens the crystal structure of co-crystal hydrate 3. Moreover, in the salt hydrate 4, the lattice water molecules form interesting R44(12) and R24(8) hydrogen bonding synthons with ADS moiety. Furthermore, the non-covalent interactions stabilizing the supramolecular assemblies of co-crystals 1-3 have been investigated using density functional theory (DFT) calculations. The molecular electrostatic potential (MEP) surfaces have been computed to rationalize the formation of the co-crystals/hydrates. In addition, the energetic features of the synthon forming O-H & ctdot;N, C-H & ctdot;O, C-H & ctdot;N, antiparallel nitrile & ctdot;nitrile and type I Cl & ctdot;Cl interactions have been analyzed using a combination of QTAIM and NCIplot computational tools.
Purpose The present study evaluated the pharmacological interaction of temozolomide (TMZ) chemotherapy with 3 potential antidepressant treatments (cannabidiol, CBD; fluoxetine, FLX; metformin, MET) in a rat glioblastoma model with the objective of finding drug combinations to better treat the depressive symptoms observed in glioblastoma patients. Methods Adult male Fisher 344 rats were inoculated with syngeneic RG2 cells into the right striatum. Then, rats received the antitumoral drug (TMZ, 25 mg/kg) vs. its vehicle (C-group), and 3 h later, treated with CBD (30 mg/kg), FLX (10 mg/kg), MET (100 mg/kg) or vehicle (Veh-group) for 5 days. The forced-swim test was used to score antidepressant-like responses. Markers of treatment response were evaluated in brain samples (hippocampal neurogenesis) of the non-tumor-infiltrated hemisphere and in blood plasma (corticosterone and immune system cells). Results The main results showed that TMZ alone and/or in combination with different types of potential antidepressants (CBD, FLX, MET) did not alter behavioral performance assessed in the forced-swim test. The absence of behavioral effects, however, is not attributed to a lack of pharmacological effects of the specific drugs. TMZ treatment decreased body weight gain, reduced the number of hippocampal neural progenitors and diminished immune plasma markers, while CBD and FLX decreased plasma corticosterone. Conclusion In conjunction, our results did not support the hypothesis of a beneficial use of antidepressants as a combinational therapy with TMZ to attenuate the most associated comorbidity of glioblastoma, depression, at least in male rats with this specific model of experimentally induced glioblastoma.