From stars to molecules: AI guides device-agnostic super-resolution imaging

A research team has developed a new super-resolution imaging method that uses artificial intelligence (AI) to enhance the quality of images obtained with various types of microscopes. This breakthrough allows overcoming the diffraction limit, a fundamental optical barrier that restricts the resolution of the finest details that can be observed. The technique, termed "device-agnostic," means it can be applied to a wide range of optical instruments, from astronomical telescopes to fluorescence microscopes, without requiring specific hardware modifications. The method is based on a deep learning algorithm trained with a diverse dataset of low-resolution images and their corresponding high-resolution versions. Once trained, the algorithm can infer and reconstruct fine details lost in the original images due to optical limitations. This is particularly useful in fields where obtaining high-resolution images is crucial but difficult, such as cell biology, materials science, and astrophysics. For instance, in microscopy, it allows visualizing subcellular structures with unprecedented clarity, while in astronomy, it could improve the sharpness of observations of distant celestial objects. The versatility of this approach lies in its ability to adapt to different optical systems and noise types, distinguishing it from other super-resolution techniques that often require very specific experimental setups or the addition of complex optical components. By being "device-agnostic," the AI acts as a universal post-processor, democratizing access to super-resolution imaging for laboratories with standard equipment. This development promises to accelerate discoveries across various disciplines by providing a powerful and flexible tool for visualizing structures at nanometer scales and beyond.

Magnetic spin and orbital Hall effects in centrosymmetric ferromagnets

Researchers have discovered the existence of magnetic spin Hall (SHE) and orbital Hall (OHE) effects in centrosymmetric ferromagnets. These effects, which involve the generation of spin or orbital angular momentum currents in response to an electric field, were previously thought to be exclusive to non-centrosymmetric materials or those with very strong spin-orbit interactions. This finding challenges conventional understanding and opens new avenues for manipulating spin and orbital currents in spintronic and orbitronic devices. Traditionally, spin and orbital Hall effects are associated with the breaking of inversion symmetry or strong spin-orbit interaction, which couples the electron's spin to its orbital motion. However, this study demonstrates that the presence of magnetic order in centrosymmetric materials is sufficient to induce these phenomena. This is because inversion symmetry is locally broken for spin-polarized electrons, even if the overall crystal structure maintains centrosymmetry. The research focused on ferromagnetic materials possessing this combination of properties. The implication of this discovery is significant for the development of new technologies. The ability to generate and manipulate spin and orbital currents in centrosymmetric ferromagnets simplifies device design, as these materials are often easier to fabricate and possess desirable magnetic properties. This could lead to advancements in magnetic memories, low-power logic, and sensors, by enabling more efficient control over spin and orbital angular momentum transport without relying on complex alloys or multilayer structures.

Classical, AI, and Quantum Algorithms for the Maximum Clique Problem

The maximum clique problem is a fundamental challenge in graph theory and computer science, with broad applications in fields such as social network analysis, bioinformatics, and optimization. It involves finding the largest complete subgraph (the "clique") within a given graph. This problem is known for its computational complexity, being NP-complete, which means no efficient algorithm is known that can solve it in polynomial time for all cases. Traditionally, various classical algorithms have been developed to tackle this problem, ranging from exhaustive methods to heuristics and approximation algorithms. However, the combinatorial explosion inherent in large graphs limits the applicability of these solutions. The emergence of artificial intelligence (AI), particularly machine learning, has opened new avenues for developing more adaptive and efficient algorithms capable of handling the complexity of modern graph data. Recently, quantum computing has emerged as a promising frontier for solving NP-complete problems. Quantum algorithms like the Quantum Approximate Optimization Algorithm (QAOA) or approaches based on Quantum Annealing are being explored to find maximum cliques. Although current quantum computers are still limited in size and reliability, they offer the potential for exponential speedup in the future, surpassing the capabilities of classical and AI algorithms for particularly difficult problem instances. Current research focuses on how to integrate and compare these three methodologies to optimize the search for maximum cliques.

Pseudo point nodal superconducting gap in spin-triplet UTe2

Researchers have discovered the existence of a pseudo point nodal superconducting gap in the compound UTe2, an exotic material exhibiting spin-triplet superconductivity. This finding is crucial for understanding the fundamental nature of unconventional superconductivity and could have significant implications for the development of fault-tolerant quantum computing. Spin-triplet superconductivity, unlike the more common spin-singlet superconductivity, involves Cooper pairs of electrons with parallel spins. This configuration is of particular interest because it is predicted to support exotic topological states, such as Majorana fermions, which are their own antiparticles and are considered promising candidates for topological qubits. UTe2 is one of the few known materials to exhibit this rare form of superconductivity. The study focused on characterizing the structure of the superconducting gap, which is the minimum energy required to break a Cooper pair. The observation of pseudo point nodal points, regions in momentum space where the gap vanishes, is a distinctive signature of certain types of unconventional superconductivity and provides vital information about the symmetry of Cooper pairs in UTe2. This knowledge is a fundamental step towards understanding and manipulating the material's topological properties.

Integrated Air-Coupled Ultrasonic Transducer with Preamplification for Structural Monitoring

Researchers have developed an integrated air-coupled ultrasonic transducer that incorporates signal preamplification. This device is designed for structural health monitoring (SHM) of materials, offering a compact and efficient solution for non-contact defect detection. The integration of preamplification within the same package as the transducer significantly improves the signal-to-noise ratio, a critical factor in applications where sound attenuation in air is considerable. The main advance lies in this transducer's ability to emit and receive ultrasonic waves through the air, eliminating the need for liquid couplants or physical contact with the structure under inspection. This is particularly useful for evaluating composite materials, metals, and other structures in environments where contact methods are unfeasible or impractical. The optimized design of the transducer, coupled with low-noise preamplification, allows for the detection of small anomalies with greater sensitivity and reliability. This development has significant implications for predictive maintenance in various industries, such as aeronautics, automotive, and civil engineering. Continuous and non-invasive monitoring of structural integrity can prevent catastrophic failures, reduce inspection costs, and extend the lifespan of critical components. The miniaturization and energy efficiency of this transducer pave the way for its implementation in autonomous SHM systems, such as drones or inspection robots.

Isogeny-based Post-Quantum Proxy Signature for IoT

Researchers have developed a new post-quantum proxy signature scheme specifically designed for Internet of Things (IoT) environments. This system is based on supersingular isogeny Diffie-Hellman (SIDH) elliptic curve cryptography, one of the most promising proposals for cryptographic security against quantum computer attacks. The proxy signature allows an original signer to delegate their signing capability to a proxy signer, who can then generate signatures on behalf of the original, a crucial functionality for managing devices and data in distributed, resource-constrained IoT networks. The main novelty of this work lies in adapting post-quantum security principles to the specificities of IoT, where devices often have significant constraints in terms of computational power, memory, and energy. The proposed scheme addresses these limitations by offering a balance between robust security and operational efficiency. The system's security is based on the computational difficulty of solving the supersingular elliptic curve isogeny problem, a problem believed to be intractable even for future quantum computers. This contrasts with current cryptographic algorithms, such as RSA or ECC, which are vulnerable to Shor's quantum algorithms. The implementation of this type of cryptography in IoT devices represents a significant step forward in ensuring data confidentiality and integrity in a future where quantum computing is a reality. The ability to securely and efficiently delegate signatures is vital for scenarios such as sensor authentication, device firmware updates, or transaction management in smart device networks. This advance contributes to building a more resilient IoT infrastructure prepared for the cryptographic challenges of the post-quantum era.

Driven-dissipative superconductivity in moiré heterostructures without attraction

Researchers have proposed a new protocol to induce superconductivity in two-dimensional moiré heterostructures, without relying on traditional attractive interactions. This method, termed driven-dissipative preparation, aims to establish superconductivity as a stationary state. The key lies in a bilayer moiré platform where the layer degree of freedom acts as a pseudospin, allowing the pseudospin structure required for pairing to be implemented through optically induced spatial operations. The scheme requires local dissipation, which naturally arises from weakly dispersive bosonic modes present in the heterostructure. In contrast, in the opposite regime of collective dissipation, the same platform exhibits an early-time superradiant burst. This approach represents an alternative to conventional superconductivity mechanisms, which generally rely on enhancing attractive interactions between electrons. The results of this research establish driven-dissipative moiré heterostructures as a promising platform for preparing superconductivity. Furthermore, the study reveals an unexpected connection between steady-state pairing and transient superradiance. This finding opens new avenues for exploring quantum phenomena and developing superconducting materials with properties controlled by light and dissipation.