arXiv subjects

Lun Li

Publications and source records attributed to Lun Li.

At least 19 recordsLinked to original sources

arXiv2026

Multi-sensor fusion for fine-guidance and milliarcsecond-level attitude estimation of balloon-borne telescope

Balloon-borne telescopes rely on fine-guidance systems to achieve milliarcsecond image stability despite residual disturbances from the balloon environment. In these systems, the Fast Steering Mirror (FSM) stabilizes the image in two focal-plane axes, but leaves systematic, field-dependent residual motion induced by boresight roll. This effect, referred to as roll leakage, becomes more important for wider fields of view. In this work, roll leakage is characterized using data from the 2023 Superpressure Balloon-borne Imaging Telescope (SuperBIT) science flight. SuperBIT is a 0.5-m near-ultraviolet to near-infrared telescope that demonstrated milliarcsecond-level image stability during its 45-night science flight. We find that passive focal-plane star-camera measurements correlate strongly with independent roll measurements across a large set of science exposures, showing that boresight roll frequently drives residual focal-plane motion. We then develop a simulation framework combining optical ray tracing, asynchronous guide-star measurements, estimation, and FSM control to study this behavior. The framework is used to compare single-star and multi-star guidance architectures under realistic flight disturbances. For the SuperBIT geometry, we find that multi-star estimation reduces average roll-induced science-field image motion by 31.8%, increasing to 77.4% for a representative geometry of GigaBIT, SuperBIT's planned larger-aperture successor. These results motivate further investigation of multi-star fine-guidance architectures for GigaBIT.

astro-ph.IM↗

arXiv2026

TaxDistill: Improving Metagenomic Taxonomic Annotation via Distilled Genomic Foundation Models

Metagenomic taxonomic annotation aims to identify the microbial origins of DNA fragments in environmental samples. Traditional methods that rely on sequence similarity are often constrained by the high microbial diversity and the incompleteness of reference databases, which has motivated the development of learning approaches such as Taxometer that perform post hoc correction to learn more informative metagenomic sequence representations. However, these methods typically rely on labels derived from similarity search tools during training, which inevitably introduces noise that can impair representation learning and degrade classification performance. To address this issue, we propose TaxDistill, a knowledge distillation framework for metagenomic classification. We introduce GenomeOcean, a 500M parameter genomic foundation model, as the teacher network to extract deep semantic features and generate soft labels based on confidence. By distilling this soft label information into a lightweight student network, TaxDistill effectively reduces the label noise introduced by initial retrieval tools. Comprehensive experiments on seven diverse CAMI2 datasets demonstrate that TaxDistill outperforms existing baselines in most scenarios. For instance, on the Gastrointestinal dataset, it improves the F1 score of MMseqs2 from 0.763 to 0.941, outperforming the Taxometer baseline. Overall, TaxDistill provides a reliable method for label correction in complex metagenomic analysis.

cs.LG↗

arXiv2026

Lensing in the Blue III: Weak Lensing Shape Catalogs of 30 Merging Galaxy Clusters

We present the weak gravitational lensing dataset from the Super-pressure Balloon-Borne Imaging Telescope (SuperBIT), which imaged 30 galaxy clusters during its 45 night flight in April to May 2023. SuperBIT is a first-of-its-kind balloon-borne imaging telescope that achieved near diffraction-limited observations in near-space conditions above 98% of the Earth's atmosphere. We use the metacalibration algorithm to obtain calibrated galaxy shapes for our target clusters and several calibration fields, enabling unbiased reconstruction of the weak-lensing signal. We employ several diagnostics throughout the pipeline, including assessments of point-spread function (PSF) modeling residuals and their impact on weak-lensing measurements, as well as tests for correlations between galaxy shapes and measured galaxy and PSF properties. To assess the multiplicative shear bias of the pipeline, we analyze a parallel set of simulated images that incorporate the real observing conditions from the flight, including measured SuperBIT PSFs, observed sky backgrounds, and detector noise, yielding a bias of $(1.1 \pm 7.8)$~per~cent.

astro-ph.CO↗

arXiv2025

PGD-based optimization of 3D bobsleigh track centerlines from 2D centerlines for simulation applications

The centerline of a bobsleigh track defines its geometry and is essential for simulation modeling. To reduce bBobsleigh training costs, leveraging the centerline of the bobsleigh track to construct a virtual environment that closely replicates real competitive settings presents a promising solution. However, publicly available centerline data are typically limited and it is imprecise to construct a training system solely based on 2-dimensional (2D) centerline. To address this practical issue, this paper proposes a method for generating a 3-dimensional (3D) track centerline based on 2D centerline data. Incorporating international track design regulations, the method formulates an optimization problem that considers total track length, height difference, slope constraints, and geometric continuity. A Projected Gradient Descent (PGD) algorithm is used to solve the optimization problem. The generated 3D centerlines are compared with real track data, and the results show that the method can reproduce realistic centerline trends from original or scaled 2D data. For the selected track segment, the relative errors in total length, height difference, and average slope are within 1.7%, 3.2% and 4.1%, respectively, for real 2D data and within 1.1%, 3.5% and 4.3% respectively for scaled data. All slope values remain within the allowable limits. Moreover, by adjusting the segmentation or modifying the weight of height difference in the cost function, various centerline styles applicable to different competitions can be generated. Under different segmentation and weight factors, the maximum errors reach up to 4.4%, 4.8%, and 9.8%, and 4.4%, 4.8%, and 10.0%, respectively. The proposed method provides a flexible and efficient tool for supporting bobsleigh track centerline design.

eess.SY↗

arXiv2025

Enhanced View Planning for Robotic Harvesting: Tackling Occlusions with Imitation Learning

In agricultural automation, inherent occlusion presents a major challenge for robotic harvesting. We propose a novel imitation learning-based viewpoint planning approach to actively adjust camera viewpoint and capture unobstructed images of the target crop. Traditional viewpoint planners and existing learning-based methods, depend on manually designed evaluation metrics or reward functions, often struggle to generalize to complex, unseen scenarios. Our method employs the Action Chunking with Transformer (ACT) algorithm to learn effective camera motion policies from expert demonstrations. This enables continuous six-degree-of-freedom (6-DoF) viewpoint adjustments that are smoother, more precise and reveal occluded targets. Extensive experiments in both simulated and real-world environments, featuring agricultural scenarios and a 6-DoF robot arm equipped with an RGB-D camera, demonstrate our method's superior success rate and efficiency, especially in complex occlusion conditions, as well as its ability to generalize across different crops without reprogramming. This study advances robotic harvesting by providing a practical "learn from demonstration" (LfD) solution to occlusion challenges, ultimately enhancing autonomous harvesting performance and productivity.

cs.RO↗

arXiv2025

Joint CSI Estimation-Feedback-Precoding via DJSCC for MU-MIMO OFDM Systems

As the number of antennas in frequency-division duplex (FDD) multiple-input multiple-output (MIMO) systems increases, acquiring channel state information (CSI) becomes increasingly challenging due to limited spectral resources and feedback overhead. In this paper, we investigate the impact of the feedback channel on CSI feedback in a multi-user MIMO orthogonal frequency-division multiplexing (OFDM) scenario, where the received downlink pilot signal is directly utilized as the source for CSI feedback in a joint design with CSI feedback and precoding. Considering the influence of the feedback channel, we propose an end-to-end joint CSI estimation-feedback-precoding network based on a deep joint source-channel coding architecture with an adaptive number of users. Experimental results demonstrate that, under the same feedback and CSI estimation overheads, the proposed joint multi-module end-to-end network achieves a higher multi-user downlink spectral efficiency than traditional algorithms based on separate architecture and partially separated artificial intelligence-based network architectures under comparable channel quality. Furthermore, compared to conventional separate architecture, the proposed network architecture with joint architecture reduces the computational burden and model storage overhead at the UE side, facilitating the deployment of low-overhead multi-module joint architectures in practice. Meanwhile, the network designed at the BS achieves user-number adaptability without increasing the number of trainable parameters, thereby reducing both model storage and distribution overhead by requiring only a single set of parameters for different numbers of users. While slightly increasing storage requirements at the base station, it reduces computational complexity and precoding design delay, effectively reducing the effects of channel aging challenges.

eess.SP↗

arXiv2024

Single-Shot 6DoF Pose and 3D Size Estimation for Robotic Strawberry Harvesting

In this study, we introduce a deep-learning approach for determining both the 6DoF pose and 3D size of strawberries, aiming to significantly augment robotic harvesting efficiency. Our model was trained on a synthetic strawberry dataset, which is automatically generated within the Ignition Gazebo simulator, with a specific focus on the inherent symmetry exhibited by strawberries. By leveraging domain randomization techniques, the model demonstrated exceptional performance, achieving an 84.77\% average precision (AP) of 3D Intersection over Union (IoU) scores on the simulated dataset. Empirical evaluations, conducted by testing our model on real-world datasets, underscored the model's viability for real-world strawberry harvesting scenarios, even though its training was based on synthetic data. The model also exhibited robust occlusion handling abilities, maintaining accurate detection capabilities even when strawberries were obscured by other strawberries or foliage. Additionally, the model showcased remarkably swift inference speeds, reaching up to 60 frames per second (FPS).

cs.RO↗

arXiv2024

SuperBIT Superpressure Flight Instrument Overview and Performance: Near diffraction-limited Astronomical Imaging from the Stratosphere

SuperBIT was a 0.5-meter near-ultraviolet to near-infrared wide-field telescope that launched on a NASA superpressure balloon into the stratosphere from New Zealand for a 45-night flight. SuperBIT acquired multi-band images of galaxy clusters to study the properties of dark matter using weak gravitational lensing. We provide an overview of the instrument and its various subsystems. We then present the instrument performance from the flight, including the telescope and image stabilization system, the optical system, the power system, and the thermal system. SuperBIT successfully met the instrument's technical requirements, achieving a telescope pointing stability of 0.34 +/- 0.10 arcseconds, a focal plane image stability of 0.055 +/- 0.027 arcseconds, and a PSF FWHM of ~ 0.35 arcseconds over 5-minute exposures throughout the 45-night flight. The telescope achieved a near-diffraction limited point-spread function in all three science bands (u, b, and g). SuperBIT served as a pathfinder to the GigaBIT observatory, which will be a 1.34-meter near-ultraviolet to near-infrared balloon-borne telescope.

astro-ph.IM↗

arXiv2024

From SuperBIT to GigaBIT: Informing next-generation balloon-borne telescope design with Fine Guidance System flight data

The Super-pressure Balloon-borne Imaging Telescope (SuperBIT) is a near-diffraction-limited 0.5m telescope that launched via NASA's super-pressure balloon technology on April 16, 2023. SuperBIT achieved precise pointing control through the use of three nested frames in conjunction with an optical Fine Guidance System (FGS), resulting in an average image stability of 0.055" over 300-second exposures. The SuperBIT FGS includes a tip-tilt fast-steering mirror that corrects for jitter on a pair of focal plane star cameras. In this paper, we leverage the empirical data from SuperBIT's successful 45-night stratospheric mission to inform the FGS design for the next-generation balloon-borne telescope. The Gigapixel Balloon-borne Imaging Telescope (GigaBIT) is designed to be a 1.35m wide-field, high resolution imaging telescope, with specifications to extend the scale and capabilities beyond those of its predecessor SuperBIT. A description and analysis of the SuperBIT FGS will be presented along with methodologies for extrapolating this data to enhance GigaBIT's FGS design and fine pointing control algorithm. We employ a systems engineering approach to outline and formalize the design constraints and specifications for GigaBIT's FGS. GigaBIT, building on the SuperBIT legacy, is set to enhance high-resolution astronomical imaging, marking a significant advancement in the field of balloon-borne telescopes.

astro-ph.IM↗

arXiv2024

When LLMs Meet Cybersecurity: A Systematic Literature Review

The rapid development of large language models (LLMs) has opened new avenues across various fields, including cybersecurity, which faces an evolving threat landscape and demand for innovative technologies. Despite initial explorations into the application of LLMs in cybersecurity, there is a lack of a comprehensive overview of this research area. This paper addresses this gap by providing a systematic literature review, covering the analysis of over 300 works, encompassing 25 LLMs and more than 10 downstream scenarios. Our comprehensive overview addresses three key research questions: the construction of cybersecurity-oriented LLMs, the application of LLMs to various cybersecurity tasks, the challenges and further research in this area. This study aims to shed light on the extensive potential of LLMs in enhancing cybersecurity practices and serve as a valuable resource for applying LLMs in this field. We also maintain and regularly update a list of practical guides on LLMs for cybersecurity at https://github.com/tmylla/Awesome-LLM4Cybersecurity.

cs.CR↗

arXiv2024

Deep Joint CSI Feedback and Multiuser Precoding for MIMO OFDM Systems

The design of precoding plays a crucial role in achieving a high downlink sum-rate in multiuser multiple-input multiple-output (MIMO) orthogonal frequency-division multiplexing (OFDM) systems. In this correspondence, we propose a deep learning based joint CSI feedback and multiuser precoding method in frequency division duplex systems, aiming at maximizing the downlink sum-rate performance in an end-to-end manner. Specifically, the eigenvectors of the CSI matrix are compressed using deep joint source-channel coding techniques. This compression method enhances the resilience of the feedback CSI information against degradation in the feedback channel. A joint multiuser precoding module and a power allocation module are designed to adjust the precoding direction and the precoding power for users based on the feedback CSI information. Experimental results demonstrate that the downlink sum-rate can be significantly improved by using the proposed method, especially in scenarios with low signal-to-noise ratio and low feedback overhead.

cs.IT↗

arXiv2024

Swin Transformer-Based CSI Feedback for Massive MIMO

For massive multiple-input multiple-output systems in the frequency division duplex (FDD) mode, accurate downlink channel state information (CSI) is required at the base station (BS). However, the increasing number of transmit antennas aggravates the feedback overhead of CSI. Recently, deep learning (DL) has shown considerable potential to reduce CSI feedback overhead. In this paper, we propose a Swin Transformer-based autoencoder network called SwinCFNet for the CSI feedback task. In particular, the proposed method can effectively capture the long-range dependence information of CSI. Moreover, we explore the impact of the number of Swin Transformer blocks and the dimension of feature channels on the performance of SwinCFNet. Experimental results show that SwinCFNet significantly outperforms other DL-based methods with comparable model sizes, especially for the outdoor scenario.

cs.IT↗

arXiv2023

Data downloaded via parachute from a NASA super-pressure balloon

In April to May 2023, the superBIT telescope was lifted to the Earth's stratosphere by a helium-filled super-pressure balloon, to acquire astronomical imaging from above (99.5% of) the Earth's atmosphere. It was launched from New Zealand then, for 40 days, circumnavigated the globe five times at a latitude 40 to 50 degrees South. Attached to the telescope were four 'DRS' (Data Recovery System) capsules containing 5 TB solid state data storage, plus a GNSS receiver, Iridium transmitter, and parachute. Data from the telescope were copied to these, and two were dropped over Argentina. They drifted 61 km horizontally while they descended 32 km, but we predicted their descent vectors within 2.4 km: in this location, the discrepancy appears irreducible below 2 km because of high speed, gusty winds and local topography. The capsules then reported their own locations to within a few metres. We recovered the capsules and successfully retrieved all of superBIT's data - despite the telescope itself being later destroyed on landing.

astro-ph.IM↗

arXiv2023

Rate Compatible LDPC Neural Decoding Network: A Multi-Task Learning Approach

Deep learning based decoding networks have shown significant improvement in decoding LDPC codes, but the neural decoders are limited by rate-matching operations such as puncturing or extending, thus needing to train multiple decoders with different code rates for a variety of channel conditions. In this correspondence, we propose a Multi-Task Learning based rate-compatible LDPC ecoding network, which utilizes the structure of raptor-like LDPC codes and can deal with multiple code rates. In the proposed network, different portions of parameters are activated to deal with distinct code rates, which leads to parameter sharing among tasks. Numerical experiments demonstrate the effectiveness of the proposed method. Training the specially designed network under multiple code rates makes the decoder compatible with multiple code rates without sacrificing frame error rate performance.

eess.SP↗

arXiv2023

Lensing in the Blue II: Estimating the Sensitivity of Stratospheric Balloons to Weak Gravitational Lensing

The Superpressure Balloon-borne Imaging Telescope (SuperBIT) is a diffraction-limited, wide-field, 0.5 m, near-infrared to near-ultraviolet observatory designed to exploit the stratosphere's space-like conditions. SuperBIT's 2023 science flight will deliver deep, blue imaging of galaxy clusters for gravitational lensing analysis. In preparation, we have developed a weak lensing measurement pipeline with modern algorithms for PSF characterization, shape measurement, and shear calibration. We validate our pipeline and forecast SuperBIT survey properties with simulated galaxy cluster observations in SuperBIT's near-UV and blue bandpasses. We predict imaging depth, galaxy number (source) density, and redshift distribution for observations in SuperBIT's three bluest filters; the effect of lensing sample selections is also considered. We find that in three hours of on-sky integration, SuperBIT can attain a depth of b = 26 mag and a total source density exceeding 40 galaxies per square arcminute. Even with the application of lensing-analysis catalog selections, we find b-band source densities between 25 and 30 galaxies per square arcminute with a median redshift of z = 1.1. Our analysis confirms SuperBIT's capability for weak gravitational lensing measurements in the blue.

astro-ph.IM↗

arXiv2022

Weak lensing in the blue: a counter-intuitive strategy for stratospheric observations

The statistical power of weak lensing measurements is principally driven by the number of high redshift galaxies whose shapes are resolved. Conventional wisdom and physical intuition suggest this is optimised by deep imaging at long (red or near IR) wavelengths, to avoid losing redshifted Balmer break and Lyman break galaxies. We use the synthetic Emission Line EL-COSMOS catalogue to simulate lensing observations using different filters, from various altitudes. Here were predict the number of exposures to achieve a target z > 0.3 source density, using off-the-shelf and custom filters. Ground-based observations are easily better at red wavelengths, as (more narrowly) are space-based observations. However, we find that SuperBIT, a diffraction-limited observatory operating in the stratosphere, should instead perform its lensing-quality observations at blue wavelengths.

astro-ph.IM↗

arXiv2022

Room Temperature Gate Tunable Non Reciprocal Charge Transport in Lattice Matched InSb/CdTe Heterostructures

The manipulation of symmetry provides an effective way to tailor the physical orders in solid-state systems. With the breaking of both the inversion and time-reversal symmetries, non-reciprocal magneto-transport may emerge in assorted non-magnetic systems to enrich spintronic physics. Here, we report the observation of the uni-directional magneto-resistance (UMR) in the lattice-matched InSb/CdTe film up to room temperature. Benefiting from the strong built-in electric field of $0.13 \mathrm{~V} \cdot \mathrm{nm}^{-1}$ in the hetero-junction region, the resulting Rashba-type spin-orbit coupling and quantum confinement warrant stable angular-dependent second-order charge current with the non-reciprocal coefficient 1-2 orders of magnitude larger than most non-centrosymmetric materials at 298 K. More importantly, this heterostructure configuration enables highly-efficient gate tuning of the rectification response in which the enhancement of the UMR amplitude by 40% is realized. Our results advocate the narrow-gap semiconductor-based hybrid system with the robust two-dimensional interfacial spin texture as a suitable platform for the pursuit of controllable chiral spin-orbit devices and applications.

cond-mat.mes-hall↗

arXiv2022

Approximation of Images via Generalized Higher Order Singular Value Decomposition over Finite-dimensional Commutative Semisimple Algebra

Low-rank approximation of images via singular value decomposition is well-received in the era of big data. However, singular value decomposition (SVD) is only for order-two data, i.e., matrices. It is necessary to flatten a higher order input into a matrix or break it into a series of order-two slices to tackle higher order data such as multispectral images and videos with the SVD. Higher order singular value decomposition (HOSVD) extends the SVD and can approximate higher order data using sums of a few rank-one components. We consider the problem of generalizing HOSVD over a finite dimensional commutative algebra. This algebra, referred to as a t-algebra, generalizes the field of complex numbers. The elements of the algebra, called t-scalars, are fix-sized arrays of complex numbers. One can generalize matrices and tensors over t-scalars and then extend many canonical matrix and tensor algorithms, including HOSVD, to obtain higher-performance versions. The generalization of HOSVD is called THOSVD. Its performance of approximating multi-way data can be further improved by an alternating algorithm. THOSVD also unifies a wide range of principal component analysis algorithms. To exploit the potential of generalized algorithms using t-scalars for approximating images, we use a pixel neighborhood strategy to convert each pixel to "deeper-order" t-scalar. Experiments on publicly available images show that the generalized algorithm over t-scalars, namely THOSVD, compares favorably with its canonical counterparts.

cs.LG↗