arXiv subjects

Zihan Li

Publications and source records attributed to Zihan Li.

At least 73 records · Page 4Linked to original sources

arXiv2025

Diffusion-Based mmWave Radar Point Cloud Enhancement Driven by Range Images

Millimeter-wave (mmWave) radar has attracted significant attention in robotics and autonomous driving. However, despite the perception stability in harsh environments, the point cloud generated by mmWave radar is relatively sparse while containing significant noise, which limits its further development. Traditional mmWave radar enhancement approaches often struggle to leverage the effectiveness of diffusion models in super-resolution, largely due to the unnatural range-azimuth heatmap (RAH) or bird's eye view (BEV) representation. To overcome this limitation, we propose a novel method that pioneers the application of fusing range images with image diffusion models, achieving accurate and dense mmWave radar point clouds that are similar to LiDAR. Benefitting from the projection that aligns with human observation, the range image representation of mmWave radar is close to natural images, allowing the knowledge from pre-trained image diffusion models to be effectively transferred, significantly improving the overall performance. Extensive evaluations on both public datasets and self-constructed datasets demonstrate that our approach provides substantial improvements, establishing a new state-of-the-art performance in generating truly three-dimensional LiDAR-like point clouds via mmWave radar. Code will be released after publication.

cs.RO↗

arXiv2025

High-pulse-energy integrated mode-locked lasers based on a Mamyshev oscillator

Ultrafast lasers have unlocked numerous advances across science and technology: they enable corneal surgery, reveal chemical reaction dynamics, and underpin optical atomic clocks. Over the past decades, extensive efforts have been devoted to developing photonic integrated circuit-based mode-locked lasers that are compact, scalable, and compatible with further on-chip functionalities. Yet, existing implementations fall short of pulse energies required for their subsequent uses in nonlinear applications. In this work, we demonstrate the first mode-locked laser that overcomes this limitation in low-loss erbium-doped silicon nitride photonic integrated circuits. The laser is based on the Mamyshev oscillator architecture, which employs alternating spectral filtering and self-phase modulation for mode-locking. It delivers a 176 MHz stream of pulses with nanojoule energy, comparable to fiber lasers and surpassing previous photonic integrated sources by more than two orders of magnitude. The output pulses exhibit excellent coherence, can be linearly compressed to 147 fs and directly drive a 1.5-octave-spanning supercontinuum in an integrated waveguide. Our work establishes a new generation of high-pulse-energy photonic integrated mode-locked lasers and paves the way for their widespread adoption.

physics.optics↗

arXiv2025

CHEMSMART: Chemistry Simulation and Modeling Automation Toolkit for High-Efficiency Computational Chemistry Workflows

CHEMSMART (Chemistry Simulation and Modeling Automation Toolkit) is an open-source, Python-based framework designed to streamline quantum chemistry workflows for homogeneous catalysis and molecular modeling. By integrating job preparation, submission, execution, results analysis, and visualization, CHEMSMART addresses the inefficiencies of manual workflow management in computational chemistry by ensuring seamless interoperability with quantum chemistry packages and cheminformatics platforms. Its modular architecture supports automated job submission and execution tasks for geometry optimization, transition state searches, thermochemical analysis, and non-covalent interaction plotting, while auxiliary scripts facilitate file conversion, data organization, and electronic structure analysis. Future developments aim to expand compatibility with additional software, incorporate QM/MM and classical MD, and align with FAIR data principles for enhanced reproducibility and data reuse. Available on GitHub, CHEMSMART empowers researchers with a robust, user-friendly platform for efficient and reproducible computational chemistry.

physics.chem-ph↗

arXiv2025

VisionUnite: A Vision-Language Foundation Model for Ophthalmology Enhanced with Clinical Knowledge

The need for improved diagnostic methods in ophthalmology is acute, especially in the underdeveloped regions with limited access to specialists and advanced equipment. Therefore, we introduce VisionUnite, a novel vision-language foundation model for ophthalmology enhanced with clinical knowledge. VisionUnite has been pretrained on an extensive dataset comprising 1.24 million image-text pairs, and further refined using our proposed MMFundus dataset, which includes 296,379 high-quality fundus image-text pairs and 889,137 simulated doctor-patient dialogue instances. Our experiments indicate that VisionUnite outperforms existing generative foundation models such as GPT-4V and Gemini Pro. It also demonstrates diagnostic capabilities comparable to junior ophthalmologists. VisionUnite performs well in various clinical scenarios including open-ended multi-disease diagnosis, clinical explanation, and patient interaction, making it a highly versatile tool for initial ophthalmic disease screening. VisionUnite can also serve as an educational aid for junior ophthalmologists, accelerating their acquisition of knowledge regarding both common and underrepresented ophthalmic conditions. VisionUnite represents a significant advancement in ophthalmology, with broad implications for diagnostics, medical education, and understanding of disease mechanisms. The source code is at https://github.com/HUANGLIZI/VisionUnite.

eess.IV↗

arXiv2025

Self-Enhanced Image Clustering with Cross-Modal Semantic Consistency

While large language-image pre-trained models like CLIP offer powerful generic features for image clustering, existing methods typically freeze the encoder. This creates a fundamental mismatch between the model's task-agnostic representations and the demands of a specific clustering task, imposing a ceiling on performance. To break this ceiling, we propose a self-enhanced framework based on cross-modal semantic consistency for efficient image clustering. Our framework first builds a strong foundation via Cross-Modal Semantic Consistency and then specializes the encoder through Self-Enhancement. In the first stage, we focus on Cross-Modal Semantic Consistency. By mining consistency between generated image-text pairs at the instance, cluster assignment, and cluster center levels, we train lightweight clustering heads to align with the rich semantics of the pre-trained model. This alignment process is bolstered by a novel method for generating higher-quality cluster centers and a dynamic balancing regularizer to ensure well-distributed assignments. In the second stage, we introduce a Self-Enhanced fine-tuning strategy. The well-aligned model from the first stage acts as a reliable pseudo-label generator. These self-generated supervisory signals are then used to feed back the efficient, joint optimization of the vision encoder and clustering heads, unlocking their full potential. Extensive experiments on six mainstream datasets show that our method outperforms existing deep clustering methods by significant margins. Notably, our ViT-B/32 model already matches or even surpasses the accuracy of state-of-the-art methods built upon the far larger ViT-L/14.

cs.CV↗

arXiv2025

Sensing and Stopping Interfering Secondary Users: Validation of an Efficient Spectrum Sharing System

We present the design and validation of Stoppable Secondary Use (StopSec), a privacy-preserving protocol with the capability to identify a secondary user (SU) causing interference to a primary user (PU) and to act quickly to stop the interference. All users are served by a database that provides a feedback mechanism from a PU to an interfering SU. We introduce a new lightweight and robust method to watermark an SU's OFDM packet. Through extensive over-the-air real-time experiments, we evaluate StopSec in terms of interference detection, identification, and stopping latency, as well as impact on SUs. We show that the watermarking method avoids negative impact to the secondary data link and is robust to real-world time-varying channels. Interfering SUs can be stopped in under 150 milliseconds, and when multiple users are simultaneously interfering, they can all be stopped. Even when the interference is 10 dB lower than the noise power, StopSec successfully stops interfering SUs within a few seconds of their appearance in the channel. StopSec can be an effective spectrum sharing protocol for cases when interference to a PU must be quickly and automatically stopped.

eess.SP↗

arXiv2025

Kling-Foley: Multimodal Diffusion Transformer for High-Quality Video-to-Audio Generation

We propose Kling-Foley, a large-scale multimodal Video-to-Audio generation model that synthesizes high-quality audio synchronized with video content. In Kling-Foley, we introduce multimodal diffusion transformers to model the interactions between video, audio, and text modalities, and combine it with a visual semantic representation module and an audio-visual synchronization module to enhance alignment capabilities. Specifically, these modules align video conditions with latent audio elements at the frame level, thereby improving semantic alignment and audio-visual synchronization. Together with text conditions, this integrated approach enables precise generation of video-matching sound effects. In addition, we propose a universal latent audio codec that can achieve high-quality modeling in various scenarios such as sound effects, speech, singing, and music. We employ a stereo rendering method that imbues synthesized audio with a spatial presence. At the same time, in order to make up for the incomplete types and annotations of the open-source benchmark, we also open-source an industrial-level benchmark Kling-Audio-Eval. Our experiments show that Kling-Foley trained with the flow matching objective achieves new audio-visual SOTA performance among public models in terms of distribution matching, semantic alignment, temporal alignment and audio quality.

eess.AS↗

arXiv2025

Integrating Large Language Models into Text Animation: An Intelligent Editing System with Inline and Chat Interaction

Text animation, a foundational element in video creation, enables efficient and cost-effective communication, thriving in advertisements, journalism, and social media. However, traditional animation workflows present significant usability barriers for non-professionals, with intricate operational procedures severely hindering creative productivity. To address this, we propose a Large Language Model (LLM)-aided text animation editing system that enables real-time intent tracking and flexible editing. The system introduces an agent-based dual-stream pipeline that integrates context-aware inline suggestions and conversational guidance as well as employs a semantic-animation mapping to facilitate LLM-driven creative intent translation. Besides, the system supports synchronized text-animation previews and parametric adjustments via unified controls to improve editing workflow. A user study evaluates the system, highlighting its ability to help non-professional users complete animation workflows while validating the pipeline. The findings encourage further exploration of integrating LLMs into a comprehensive video creation workflow.

cs.HC↗

arXiv2025

FIC-TSC: Learning Time Series Classification with Fisher Information Constraint

Analyzing time series data is crucial to a wide spectrum of applications, including economics, online marketplaces, and human healthcare. In particular, time series classification plays an indispensable role in segmenting different phases in stock markets, predicting customer behavior, and classifying worker actions and engagement levels. These aspects contribute significantly to the advancement of automated decision-making and system optimization in real-world applications. However, there is a large consensus that time series data often suffers from domain shifts between training and test sets, which dramatically degrades the classification performance. Despite the success of (reversible) instance normalization in handling the domain shifts for time series regression tasks, its performance in classification is unsatisfactory. In this paper, we propose \textit{FIC-TSC}, a training framework for time series classification that leverages Fisher information as the constraint. We theoretically and empirically show this is an efficient and effective solution to guide the model converge toward flatter minima, which enhances its generalizability to distribution shifts. We rigorously evaluate our method on 30 UEA multivariate and 85 UCR univariate datasets. Our empirical results demonstrate the superiority of the proposed method over 14 recent state-of-the-art methods.

cs.LG↗

arXiv2025

Copper Damascene Process-Based High-Performance Thin Film Lithium Tantalate Modulators

Interfacing electrical and optical systems is a ubiquitous requirement for modern networks. Competitive footprint, efficiency, and bandwidth figures have propelled interest in deploying integrated electro-optic modulators based on Pockels materials for such tasks. Due to its wide usage in legacy bulk electro-optic modulators, and triggered by the availability of 'on insulator' wafers, lithium niobate based devices have seen major advances. Recently, even more favorable properties have been demonstrated in lithium tantalate based devices, featuring similar Pockels effect, but exhibiting lower bias drift and lower birefringence, while equally benefiting from existing volume usage in wireless RF filters. Despite major progress in integrated modulators, these newly emerged ferro-electrical modulators cannot be integrated tightly with electronics yet using standardized processes, such as flip-chip bonding. Here, we overcome this bottleneck by incorporating the copper Damascene process in the fabrication of integrated lithium tantalate modulators. We demonstrate modulators featuring microwave losses that are ~10% lower than in designs relying on conventional gold electrodes. Our results allow us to reach data transmission figures on par with those of electro-optic modulators fabricated with other low-resistivity, yet less common, metals. Specifically, our fabricated modulators are able to achieve data transmission rates of 416 and 540 Gbit/s while preserving bit error ratios below the 25% SD-FEC threshold in a PAM4 and PAM8 transmission schemes, respectively. Together with the commercial availability of lithium tantalate as a Pockels material, our results open a path towards scalable and direct chip-on-wafer embedding of EO modulators with micro-electronic integrated circuits.

physics.optics↗

arXiv2025

STPNet: Scale-aware Text Prompt Network for Medical Image Segmentation

Accurate segmentation of lesions plays a critical role in medical image analysis and diagnosis. Traditional segmentation approaches that rely solely on visual features often struggle with the inherent uncertainty in lesion distribution and size. To address these issues, we propose STPNet, a Scale-aware Text Prompt Network that leverages vision-language modeling to enhance medical image segmentation. Our approach utilizes multi-scale textual descriptions to guide lesion localization and employs retrieval-segmentation joint learning to bridge the semantic gap between visual and linguistic modalities. Crucially, STPNet retrieves relevant textual information from a specialized medical text repository during training, eliminating the need for text input during inference while retaining the benefits of cross-modal learning. We evaluate STPNet on three datasets: COVID-Xray, COVID-CT, and Kvasir-SEG. Experimental results show that our vision-language approach outperforms state-of-the-art segmentation methods, demonstrating the effectiveness of incorporating textual semantic knowledge into medical image analysis. The code has been made publicly on https://github.com/HUANGLIZI/STPNet.

eess.IV↗

arXiv2025

Power Domain Sparse Dimensional Constellation Multiple Access (PD-SDCMA): A Novel PD-NOMA for More Access Users

With the advent of the 6G mobile communication network era, the existing non-orthogonal multiple-access (NOMA) technology faces the challenge of high successive interference in multi-user scenarios, which limits its ability to support more user access. To address this, this paper proposes a novel power-domain sparse-dimensional constellation multiple-access scheme (PD-SDCMA). Through the signal space dimension selection strategy (S2D-strategy), this scheme sparsely superposes low-dimensional constellations onto high-dimensional signal spaces, and reduces the high-order interference caused by SC by taking advantage of the non-correlation between dimensions. Specifically, PD-SDCMA reduces the successive interference between users by sparsifying the dimension allocation of constellation superposition and designs a sparse superposition method based on the theory of vector space signal representation. Simulation results show that, under the AWGN channel, PD-SDCMA significantly outperforms the traditional PD-NOMA in terms of the number of supported users under QPSK and 16QAM modulations, and also has better BER performance. This paper provides a new solution for efficient spectrum utilization in future scenarios with large-scale user access.

cs.ET↗

arXiv2025

An Intra- and Cross-frame Topological Consistency Scheme for Semi-supervised Atherosclerotic Coronary Plaque Segmentation

Enhancing the precision of segmenting coronary atherosclerotic plaques from CT Angiography (CTA) images is pivotal for advanced Coronary Atherosclerosis Analysis (CAA), which distinctively relies on the analysis of vessel cross-section images reconstructed via Curved Planar Reformation. This task presents significant challenges due to the indistinct boundaries and structures of plaques and blood vessels, leading to the inadequate performance of current deep learning models, compounded by the inherent difficulty in annotating such complex data. To address these issues, we propose a novel dual-consistency semi-supervised framework that integrates Intra-frame Topological Consistency (ITC) and Cross-frame Topological Consistency (CTC) to leverage labeled and unlabeled data. ITC employs a dual-task network for simultaneous segmentation mask and Skeleton-aware Distance Transform (SDT) prediction, achieving similar prediction of topology structure through consistency constraint without additional annotations. Meanwhile, CTC utilizes an unsupervised estimator for analyzing pixel flow between skeletons and boundaries of adjacent frames, ensuring spatial continuity. Experiments on two CTA datasets show that our method surpasses existing semi-supervised methods and approaches the performance of supervised methods on CAA. In addition, our method also performs better than other methods on the ACDC dataset, demonstrating its generalization.

eess.IV↗

arXiv2024

LLM-RG4: Flexible and Factual Radiology Report Generation across Diverse Input Contexts

Drafting radiology reports is a complex task requiring flexibility, where radiologists tail content to available information and particular clinical demands. However, most current radiology report generation (RRG) models are constrained to a fixed task paradigm, such as predicting the full ``finding'' section from a single image, inherently involving a mismatch between inputs and outputs. The trained models lack the flexibility for diverse inputs and could generate harmful, input-agnostic hallucinations. To bridge the gap between current RRG models and the clinical demands in practice, we first develop a data generation pipeline to create a new MIMIC-RG4 dataset, which considers four common radiology report drafting scenarios and has perfectly corresponded input and output. Secondly, we propose a novel large language model (LLM) based RRG framework, namely LLM-RG4, which utilizes LLM's flexible instruction-following capabilities and extensive general knowledge. We further develop an adaptive token fusion module that offers flexibility to handle diverse scenarios with different input combinations, while minimizing the additional computational burden associated with increased input volumes. Besides, we propose a token-level loss weighting strategy to direct the model's attention towards positive and uncertain descriptions. Experimental results demonstrate that LLM-RG4 achieves state-of-the-art performance in both clinical efficiency and natural language generation on the MIMIC-RG4 and MIMIC-CXR datasets. We quantitatively demonstrate that our model has minimal input-agnostic hallucinations, whereas current open-source models commonly suffer from this problem.

cs.CL↗

arXiv2024

Ultrabroadband thin-film lithium tantalate modulator for high-speed communications

The continuous growth of global data traffic over the past three decades, along with advances in disaggregated computing architectures, presents significant challenges for optical transceivers in communication networks and high-performance computing systems. Specifically, there is a growing need to significantly increase data rates while reducing energy consumption and cost. High-performance optical modulators based on materials such as InP, thin-film lithium niobate (LiNbO3), or plasmonics have been developed, with LiNbO3 excelling in high-speed and low-voltage modulation. Nonetheless, the widespread industrial adoption of thin film LiNbO3 remains compounded by the rather high cost of the underlying 'on insulator' substrates -- in sharp contrast to silicon photonics, which can benefit from strong synergies with high-volume applications in conventional microelectronics. Here, we demonstrate an integrated 110 GHz modulator using thin-film lithium tantalate (LiTaO3) -- a material platform that is already commercially used for millimeter-wave filters and that can hence build upon technological and economic synergies with existing high-volume applications to offer scalable low-cost manufacturing. We show that the LiTaO3 photonic integrated circuit based modulator can support 176 GBd PAM8 transmission at net data rates exceeding 400 Gbit/s. Moreover, we show that using silver electrodes can reduce microwave losses compared to previously employed gold electrodes. Our demonstration positions LiTaO3 modulator as a novel and highly promising integration platform for next-generation high-speed, energy-efficient, and cost-effective transceivers.

physics.optics↗

arXiv2024

Integrated Triply Resonant Electro-Optic Frequency Comb in Lithium Tantalate

Integrated frequency comb generators based on Kerr parametric oscillation have led to chip-scale, gigahertz-spaced combs with new applications spanning hyperscale telecommunications, low-noise microwave synthesis, LiDAR, and astrophysical spectrometer calibration. Recent progress in lithium niobate (LN) photonic integrated circuits (PICs) has resulted in chip-scale electro-optic (EO) frequency combs, offering precise comb-line positioning and simple operation without relying on the formation of dissipative Kerr solitons. However, current integrated EO combs face limited spectral coverage due to the large microwave power required to drive the non-resonant capacitive electrodes and the strong intrinsic birefringence of Lithium Niobate. Here, we overcome both challenges with an integrated triply resonant architecture, combining monolithic microwave integrated circuits (MMICs) with PICs based on the recently emerged thin-film lithium tantalate. With resonantly enhanced EO interaction and reduced birefringence in Lithium Tantalate, we achieve a four-fold comb span extension and a 16-fold power reduction compared to the conventional non-resonant microwave design. Driven by a hybrid-integrated laser diode, the comb spans over 450nm (60THz) with >2000 lines, and the generator fits within a compact 1cm^2 footprint. We additionally observe that the strong EO coupling leads to an increased comb existence range approaching the full free spectral range of the optical microresonator. The ultra-broadband comb generator, combined with detuning-agnostic operation, could advance chip-scale spectrometry and ultra-low-noise millimeter wave synthesis and unlock octave-spanning EO combs. The methodology of co-designing microwave and optical resonators can be extended to a wide range of integrated electro-optics applications.

physics.optics↗

arXiv2024

Fundamental charge noise in electro-optic photonic integrated circuits

Understanding thermodynamical measurement noise is of central importance for electrical and optical precision measurements from mass-fabricated semiconductor sensors, where the Brownian motion of charge carriers poses limits, to optical reference cavities for atomic clocks or gravitational wave detection, which are limited by thermorefractive and thermoelastic noise due to the transduction of temperature fluctuations to the refractive index and length fluctuations. Here, we discover that unexpectedly charge carrier density fluctuations give rise to a novel noise process in recently emerged electro-optic photonic integrated circuits. We show that Lithium Niobate and Lithium Tantalate photonic integrated microresonators exhibit an unexpected Flicker type (i.e. $1/f^{1.2}$) scaling in their noise properties, significantly deviating from the well-established thermorefractive noise theory. We show that this noise is consistent with thermodynamical charge noise, which leads to electrical field fluctuations that are transduced via the strong Pockels effects of electro-optic materials. Our results establish electrical Johnson-Nyquist noise as the fundamental limitation for Pockels integrated photonics, crucial for determining performance limits for both classical and quantum devices, ranging from ultra-fast tunable and low-noise lasers, Pockels soliton microcombs, to quantum transduction, squeezed light or entangled photon-pair generation. Equally, this observation offers optical methods to probe mesoscopic charge fluctuations with exceptional precision.

physics.optics↗

arXiv2024

Are Large Language Models Possible to Conduct Cognitive Behavioral Therapy?

In contemporary society, the issue of psychological health has become increasingly prominent, characterized by the diversification, complexity, and universality of mental disorders. Cognitive Behavioral Therapy (CBT), currently the most influential and clinically effective psychological treatment method with no side effects, has limited coverage and poor quality in most countries. In recent years, researches on the recognition and intervention of emotional disorders using large language models (LLMs) have been validated, providing new possibilities for psychological assistance therapy. However, are LLMs truly possible to conduct cognitive behavioral therapy? Many concerns have been raised by mental health experts regarding the use of LLMs for therapy. Seeking to answer this question, we collected real CBT corpus from online video websites, designed and conducted a targeted automatic evaluation framework involving the evaluation of emotion tendency of generated text, structured dialogue pattern and proactive inquiry ability. For emotion tendency, we calculate the emotion tendency score of the CBT dialogue text generated by each model. For structured dialogue pattern, we use a diverse range of automatic evaluation metrics to compare speaking style, the ability to maintain consistency of topic and the use of technology in CBT between different models . As for inquiring to guide the patient, we utilize PQA (Proactive Questioning Ability) metric. We also evaluated the CBT ability of the LLM after integrating a CBT knowledge base to explore the help of introducing additional knowledge to enhance the model's CBT counseling ability. Four LLM variants with excellent performance on natural language processing are evaluated, and the experimental result shows the great potential of LLMs in psychological counseling realm, especially after combining with other technological means.

cs.CL↗