A Generative Approach to Person Reidentification.

Person Re-identification is the task of recognizing comparable subjects across a network of nonoverlapping cameras. This is typically achieved by extracting from the source image a vector of characteristic features of the specific person captured by the camera. Learning a good set of robust, invariant and discriminative features is a complex task, often leveraging contrastive learning. In this article, we explore a different approach, learning the representation of an individual as the conditioning information required to generate images of the specific person starting from random noise. In this way we decouple the identity of the individual from any other information relative to a specific instance (pose, background, etc.), allowing interesting transformations from one identity to another. As generative models, we use the recent diffusion models that have already proven their sensibility to conditioning in many different contexts. The results presented in this article serve as a proof-of-concept. While our current performance on common benchmarks is lower than state-of-the-art techniques, the approach is intriguing and rich of innovative insights, suggesting a wide range of potential improvements along various lines of investigation.

High-quality nanocavities through multimodal confinement of hyperbolic polaritons in hexagonal boron nitride.

Compressing light into nanocavities substantially enhances light-matter interactions, which has been a major driver for nanostructured materials research. However, extreme confinement generally comes at the cost of absorption and low resonator quality factors. Here we suggest an alternative optical multimodal confinement mechanism, unlocking the potential of hyperbolic phonon polaritons in isotopically pure hexagonal boron nitride. We produce deep-subwavelength cavities and demonstrate several orders of magnitude improvement in confinement, with estimated Purcell factors exceeding 108 and quality factors in the 50-480 range, values approaching the intrinsic quality factor of hexagonal boron nitride polaritons. Intriguingly, the quality factors we obtain exceed the maximum predicted by impedance-mismatch considerations, indicating that confinement is boosted by higher-order modes. We expect that our multimodal approach to nanoscale polariton manipulation will have far-reaching implications for ultrastrong light-matter interactions, mid-infrared nonlinear optics and nanoscale sensors.

Transverse Hypercrystals Formed by Periodically Modulated Phonon Polaritons.

Photonic crystals and metamaterials are two overarching paradigms for manipulating light. By combining these approaches, hypercrystals can be created, which are hyperbolic dispersion metamaterials that undergo periodic modulation and mix photonic-crystal-like aspects with hyperbolic dispersion physics. Despite several attempts, there has been limited experimental realization of hypercrystals due to technical and design constraints. In this work, hypercrystals with nanoscale lattice constants ranging from 25 to 160 nm were created. The Bloch modes of these crystals were then measured directly using scattering near-field microscopy. The dispersion of the Bloch modes was extracted from the frequency dependence of the Bloch modes, revealing a clear switch from positive to negative group velocity. Furthermore, spectral features specific to hypercrystals were observed in the form of sharp density of states peaks, which are a result of intermodal coupling and should not appear in ordinary polaritonic crystals with an equivalent geometry. These findings are in agreement with theoretical predictions that even simple lattices can exhibit a rich hypercrystal bandstructure. This work is of both fundamental and practical interest, providing insight into nanoscale light-matter interactions and the potential to manipulate the optical density of states.

2D-3D integration of hexagonal boron nitride and a high-κ dielectric for ultrafast graphene-based electro-absorption modulators.

Electro-absorption (EA) waveguide-coupled modulators are essential building blocks for on-chip optical communications. Compared to state-of-the-art silicon (Si) devices, graphene-based EA modulators promise smaller footprints, larger temperature stability, cost-effective integration and high speeds. However, combining high speed and large modulation efficiencies in a single graphene-based device has remained elusive so far. In this work, we overcome this fundamental trade-off by demonstrating the 2D-3D dielectric integration in a high-quality encapsulated graphene device. We integrated hafnium oxide (HfO2) and two-dimensional hexagonal boron nitride (hBN) within the insulating section of a double-layer (DL) graphene EA modulator. This combination of materials allows for a high-quality modulator device with high performances: a ~39 GHz bandwidth (BW) with a three-fold increase in modulation efficiency compared to previously reported high-speed modulators. This 2D-3D dielectric integration paves the way to a plethora of electronic and opto-electronic devices with enhanced performance and stability, while expanding the freedom for new device designs.

More fluorous surface modifier makes it less oleophobic: fluorinated siloxane copolymer/PDMS coatings.

A copolyacrylate with semifluorinated and polydimethylsiloxane side chains (D5-3) was used as a surface modifier for a condensation-cured PDMS coating. The decyl fluorous group is represented by "D"; "5" is a 5 kDa silicone, and "3" is the mole ratio of fluorous to silicone side chains. Wetting behavior was assessed by dynamic contact angle (DCA) analysis using isopropanol, which differentiates silicone and fluorous wetting behavior. Interestingly, a maximum in surface oleophobicity was found at low D5-3 concentration (0.4 wt %). Higher concentrations result in decreased oleophobicity, as reflected in decreased contact angles. To understand this unexpected observation, dynamic light scattering (DLS) studies were initiated on a model system consisting of hydroxyl-terminated PDMS (18 kDa) containing varying amounts of D5-3. DLS revealed D5-3 aggregation to be a function of temperature and concentration. A model is proposed by which D5-3 surface concentration is depleted via phase separation favoring D5-3 aggregation at concentrations >0.4 wt %, that is, the cmc. This model suggests increasing aggregate/micelle concentrations at increased D5-3 concentration. Bulk morphologies studied by scanning electron microscopy (SEM) and atomic force microscopy (AFM) support this model by showing increased aggregate concentrations with increased D5-3 > 0.4 wt %.

Fluorinated/siloxane copolymer blends for fouling release: chemical characterisation and biological evaluation with algae and barnacles.

Fouling-release coatings were prepared from blends of a fluorinated/siloxane copolymer with a poly(dimethyl siloxane) (PDMS) matrix in order to couple the low modulus character of PDMS with the low surface tension typical for fluorinated polymers. The content of the surface-active copolymer was varied in the blend over a broad range (0.15-10 wt % with respect to PDMS). X-ray photoelectron spectroscopy depth profiling analyses were performed on the coatings to establish the distribution of specific chemical constituents throughout the coatings, and proved enrichment in fluorine of the outermost layers of the coating surface. Addition of the fluorinated/siloxane copolymer to the PDMS matrix resulted in a concentration-dependent decrease in settlement of barnacle, Balanus amphitrite, cyprids. The release of young plants of Ulva, a soft fouling species, and young barnacles showed that adhesion strength on the fluorinated/siloxane copolymer was significantly lower than the siloxane control. However, differences in adhesion strength were not directly correlated with the concentration of copolymer in the blends.