BZPEERReading, trust and collaboration for researchDiscover
Workspaces
Network
Opportunities
Account
Researcher profile
David Yang contributes to research discovery and scholarly infrastructure.
Trust snapshot
Actions
Identity and collaboration
Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.
Log in to claimDirect collaboration
Claim this author entity first to unlock direct invitations.
Research graph
Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.
BZPEER is loading the nearby papers, people, topics and institutions for this page.
Published work
We prove the Frenkel-Gaitsgory localization conjecture describing regular Kac-Moody representations at critical level via eigensheaves on the affine Grassmannian using categorical Moy-Prasad theory. This extends previous work of the authors.
Corrosion is a major concern for many industries, as corrosive environments can induce structural and morphological changes that lead to material dissolution and accelerate material failure. The progression of corrosion depends on nanoscale morphology, stress, and defects present. Experimentally monitoring this complex interplay is challenging. Here we implement in situ Bragg coherent X-ray diffraction imaging (BCDI) to probe the dissolution of a Co-Fe alloy microcrystal exposed to hydrochloric acid (HCl). By measuring five Bragg reflections from a single isolated microcrystal at ambient conditions, we compare the full three-dimensional (3D) strain state before corrosion and the strain along the [111] direction throughout the corrosion process. We find that the strained surface layer of the crystal dissolves to leave a progressively less strained surface. Interestingly, the average strain closer to the centre of the crystal increases during the corrosion process. We determine the localised corrosion rate from BCDI data, revealing the preferential dissolution of facets more exposed to the acid stream, highlighting an experimental geometry effect. These results bring new perspectives to understanding the interplay between crystal strain, morphology, and corrosion; a prerequisite for the design of more corrosion-resistant materials.
Image captioning is a longstanding problem in the field of computer vision and natural language processing. To date, researchers have produced impressive state-of-the-art performance in the age of deep learning. Most of these state-of-the-art, however, requires large volume of annotated image-caption pairs in order to train their models. When given an image dataset of interests, practitioner needs to annotate the caption for each image in the training set and this process needs to happen for each newly collected image dataset. In this paper, we explore the task of unsupervised image captioning which utilizes unpaired images and texts to train the model so that the texts can come from different sources than the images. A main school of research on this topic that has been shown to be effective is to construct pairs from the images and texts in the training set according to their overlap of objects. Unlike in the supervised setting, these constructed pairings are however not guaranteed to have fully overlapping set of objects. Our work in this paper overcomes this by harvesting objects corresponding to a given sentence from the training set, even if they don't belong to the same image. When used as input to a transformer, such mixture of objects enables larger if not full object coverage, and when supervised by the corresponding sentence, produced results that outperform current state of the art unsupervised methods by a significant margin. Building upon this finding, we further show that (1) additional information on relationship between objects and attributes of objects also helps in boosting performance; and (2) our method also extends well to non-English image captioning, which usually suffers from a scarcer level of annotations. Our findings are supported by strong empirical results. Our code is available at https://github.com/zihangm/obj-centric-unsup-caption.
Focused ion beam (FIB) techniques are commonly used to machine, analyse and image materials at the micro- and nanoscale. However, FIB modifies the integrity of the sample by creating defects that cause lattice distortions. Methods have been developed to reduce FIB-induced strain, however these protocols need to be evaluated for their effectiveness. Here we use non-destructive Bragg coherent X-ray diffraction imaging to study the in situ annealing of FIB-milled gold microcrystals. We simultaneously measure two non-collinear reflections for two different crystals during a single annealing cycle, demonstrating the ability to reliably track the location of multiple Bragg peaks during thermal annealing. The thermal lattice expansion of each crystal is used to calculate the local temperature. This is compared to thermocouple readings, which are shown to be substantially affected by thermal resistance. To evaluate the annealing process, we analyse each reflection by considering facet area evolution, cross-correlation maps of displacement field and binarised morphology, and average strain plots. The crystal's strain and morphology evolve with increasing temperature, which is likely to be caused by the diffusion of gallium in gold below ~280°C and the self-diffusion of gold above ~280°C. The majority of FIB-induced strains are removed by 380-410\degree C, depending on which reflection is being considered. Our observations highlight the importance of measuring multiple reflections to unambiguously interpret material behaviour.
Bragg coherent X-ray diffraction imaging (BCDI) is a non-destructive, lensless method for 3D-resolved, nanoscale strain imaging in micro-crystals. A challenge, particularly for new users of the technique, is accurate mapping of experimental data, collected in the detector reciprocal space coordinate frame, to more convenient orthogonal coordinates, e.g. attached to the sample. This is particularly the case since different coordinate conventions are used at every BCDI beamline. The reconstruction algorithms and mapping scripts composed for individual beamlines are not readily interchangeable. To overcome this, a BCDI experiment simulation with a plugin script that converts all beamline angles to a universal, right-handed coordinate frame is introduced, making it possible to condense any beamline geometry into three rotation matrices. The simulation translates a user-specified 3D complex object to different BCDI-related coordinate frames. It also allows the generation of synthetic coherent diffraction data that can be inserted into any BCDI reconstruction algorithm to reconstruct the original user-specified object. Scripts are provided to map from sample space to detector conjugated space, detector conjugated space to sample space and detector conjugated space to detector conjugated space for a different reflection. This provides the reader with the basis for a flexible simulation tool kit that is easily adapted to different geometries. It is anticipated that this will find use in the generation of tailor-made supports for phasing of challenging data and exploration of novel geometries or data collection modalities.
Human Body Communication (HBC) has come up as a promising alternative to traditional radio frequency (RF) Wireless Body Area Network (WBAN) technologies. This is essentially due to HBC providing a broadband communication channel with enhanced signal security in the physical layer due to lower radiation from the human body as compared to its RF counterparts. An in-depth understanding of the mechanism for the channel loss variability and associated biophysical model needs to be developed before EQS-HBC can be used more frequently in WBAN consumer and medical applications. Biophysical models characterizing the human body as a communication channel didn't exist in literature for a long time. Recent developments have shown models that capture the channel response for fixed transmitter and receiver positions on the human body. These biophysical models do not capture the variability in the HBC channel for varying positions of the devices with respect to the human body. In this study, we provide a detailed analysis of the change in path loss in a capacitive-HBC channel in the electroquasistatic (EQS) domain. Causes of channel loss variability namely: inter-device coupling and effects of fringe fields due to body's shadowing effects are investigated. FEM based simulation results are used to analyze the channel response of human body for different positions and sizes of the device which are further verified using measurement results to validate the developed biophysical model. Using the bio-physical model, we develop a closed form equation for the path loss in a capacitive HBC channel which is then analyzed as a function of the geometric properties of the device and the position with respect to the human body which will help pave the path towards future EQSHBC WBAN design.