arXiv subjects

Lin Li

Publications and source records attributed to Lin Li.

At least 127 records · Page 7Linked to original sources

arXiv2025

Transplant-Ready? Evaluating AI Lung Segmentation Models in Candidates with Severe Lung Disease

This study evaluates publicly available deep-learning based lung segmentation models in transplant-eligible patients to determine their performance across disease severity levels, pathology categories, and lung sides, and to identify limitations impacting their use in preoperative planning in lung transplantation. This retrospective study included 32 patients who underwent chest CT scans at Duke University Health System between 2017 and 2019 (total of 3,645 2D axial slices). Patients with standard axial CT scans were selected based on the presence of two or more lung pathologies of varying severity. Lung segmentation was performed using three previously developed deep learning models: Unet-R231, TotalSegmentator, MedSAM. Performance was assessed using quantitative metrics (volumetric similarity, Dice similarity coefficient, Hausdorff distance) and a qualitative measure (four-point clinical acceptability scale). Unet-R231 consistently outperformed TotalSegmentator and MedSAM in general, for different severity levels, and pathology categories (p<0.05). All models showed significant performance declines from mild to moderate-to-severe cases, particularly in volumetric similarity (p<0.05), without significant differences among lung sides or pathology types. Unet-R231 provided the most accurate automated lung segmentation among evaluated models with TotalSegmentator being a close second, though their performance declined significantly in moderate-to-severe cases, emphasizing the need for specialized model fine-tuning in severe pathology contexts.

cs.CV↗

arXiv2025

RED: Unleashing Token-Level Rewards from Holistic Feedback via Reward Redistribution

Reinforcement learning from human feedback (RLHF) offers a promising approach to aligning large language models (LLMs) with human preferences. Typically, a reward model is trained or supplied to act as a proxy for humans in evaluating generated responses during the reinforcement training phase. However, current reward models operate as sequence-to-one models, allocating a single, sparse, and delayed reward to an entire output sequence. This approach may overlook the significant contributions of individual tokens toward the desired outcome. To this end, we propose a more fine-grained, token-level guidance approach for RL training. Specifically, we introduce RED, a novel reward redistribition method that evaluates and assigns specific credit to each token using an off-the-shelf reward model. Utilizing these fine-grained rewards enhances the model's understanding of language nuances, leading to more precise performance improvements. Notably, our method does not require modifying the reward model or introducing additional training steps, thereby incurring minimal computational costs. Experimental results across diverse datasets and tasks demonstrate the superiority of our approach.

cs.CL↗

arXiv2025

Energy Landscapes Enable Reliable Abstention in Retrieval-Augmented Large Language Models for Healthcare

Reliable abstention is critical for retrieval-augmented generation (RAG) systems, particularly in safety-critical domains such as women's health, where incorrect answers can lead to harm. We present an energy-based model (EBM) that learns a smooth energy landscape over a dense semantic corpus of 2.6M guideline-derived questions, enabling the system to decide when to generate or abstain. We benchmark the EBM against a calibrated softmax baseline and a k-nearest neighbour (kNN) density heuristic across both easy and hard abstention splits, where hard cases are semantically challenging near-distribution queries. The EBM achieves superior abstention performance abstention on semantically hard cases, reaching AUROC 0.961 versus 0.950 for softmax, while also reducing FPR@95 (0.235 vs 0.331). On easy negatives, performance is comparable across methods, but the EBM's advantage becomes most pronounced in safety-critical hard distributions. A comprehensive ablation with controlled negative sampling and fair data exposure shows that robustness stems primarily from the energy scoring head, while the inclusion or exclusion of specific negative types (hard, easy, mixed) sharpens decision boundaries but is not essential for generalisation to hard cases. These results demonstrate that energy-based abstention scoring offers a more reliable confidence signal than probability-based softmax confidence, providing a scalable and interpretable foundation for safe RAG systems.

cs.CL↗

arXiv2025

XMUspeech Systems for the ASVspoof 5 Challenge

In this paper, we present our submitted XMUspeech systems to the speech deepfake detection track of the ASVspoof 5 Challenge. Compared to previous challenges, the audio duration in ASVspoof 5 database has significantly increased. And we observed that merely adjusting the input audio length can substantially improve system performance. To capture artifacts at multiple levels, we explored the performance of AASIST, HM-Conformer, Hubert, and Wav2vec2 with various input features and loss functions. Specifically, in order to obtain artifact-related information, we trained self-supervised models on the dataset containing spoofing utterances as the feature extractors. And we applied an adaptive multi-scale feature fusion (AMFF) method to integrate features from multiple Transformer layers with the hand-crafted feature to enhance the detection capability. In addition, we conducted extensive experiments on one-class loss functions and provided optimized configurations to better align with the anti-spoofing task. Our fusion system achieved a minDCF of 0.4783 and an EER of 20.45% in the closed condition, and a minDCF of 0.2245 and an EER of 9.36% in the open condition.

cs.SD↗

arXiv2025

VoxHammer: Training-Free Precise and Coherent 3D Editing in Native 3D Space

3D local editing of specified regions is crucial for game industry and robot interaction. Recent methods typically edit rendered multi-view images and then reconstruct 3D models, but they face challenges in precisely preserving unedited regions and overall coherence. Inspired by structured 3D generative models, we propose VoxHammer, a novel training-free approach that performs precise and coherent editing in 3D latent space. Given a 3D model, VoxHammer first predicts its inversion trajectory and obtains its inverted latents and key-value tokens at each timestep. Subsequently, in the denoising and editing phase, we replace the denoising features of preserved regions with the corresponding inverted latents and cached key-value tokens. By retaining these contextual features, this approach ensures consistent reconstruction of preserved areas and coherent integration of edited parts. To evaluate the consistency of preserved regions, we constructed Edit3D-Bench, a human-annotated dataset comprising hundreds of samples, each with carefully labeled 3D editing regions. Experiments demonstrate that VoxHammer significantly outperforms existing methods in terms of both 3D consistency of preserved regions and overall quality. Our method holds promise for synthesizing high-quality edited paired data, thereby laying the data foundation for in-context 3D generation. See our project page at https://huanngzh.github.io/VoxHammer-Page/.

cs.CV↗

arXiv2025

Enhanced superconductivity in PbTe-In hybrids

We report the realization of epitaxial indium thin films on PbTe nanowires. The film is continuous and forms an atomically sharp interface with PbTe. Tunneling devices reveal a hard superconducting gap. The gap size, 1.08 to 1.18 meV, is twice as large as bulk indium's ($\sim$ 0.5 meV), due to the presence of PbTe. A similar enhancement is also observed in the critical temperature of In on a PbTe substrate. Subgap peaks appear at finite magnetic fields. The effective $g$-factor (15-45) is notably enhanced compared to bare PbTe wires ($<$ 10) due to the presence of In, differing from Al-hybrids. Josephson devices exhibit gate-tunable supercurrents. The PbTe-In hybrid enhances the properties of both, the superconductivity of In and g-factors of PbTe, and thus may enable exotic phases of matter such as topological superconductivity.

cond-mat.mes-hall↗

arXiv2025

Synchrosqueezed X-Ray Wavelet-Chirplet Transform for Accurate Chirp Rate Estimation and Retrieval of Modes from Multicomponent Signals with Crossover Instantaneous Frequencies

Recent advances in the chirplet transform and wavelet-chirplet transform (WCT) have enabled the estimation of instantaneous frequencies (IFs) and chirprates, as well as mode retrieval from multicomponent signals with crossover IF curves. However, chirprate estimation via these approaches remains less accurate than IF estimation, primarily due to the slow decay of the chirplet transform or WCT along the chirprate direction. To address this, the synchrosqueezed chirplet transform (SCT) and multiple SCT methods were proposed, achieving moderate improvements in IF and chirprate estimation accuracy. Nevertheless, a novel approach is still needed to enhance the transform's decay along the chirprate direction. This paper introduces an X-ray transform-based wavelet-chirprate transform, termed the X-ray wavelet-chirplet transform (XWCT), which exhibits superior decay along the chirprate direction compared to the WCT. Furthermore, third-order synchrosqueezed variants of the WCT and XWCT are developed to yield sharp time-frequency-chirprate representations of signals. Experimental results demonstrate that the XWCT achieves significantly faster decay along the chirprate axis, while the third-order synchrosqueezed XWCT enables accurate IF and chirprate estimation, as well as mode retrieval, without requiring multiple synchrosqueezing operations.

eess.SP↗

arXiv2025

A Node-Aware Dynamic Quantization Approach for Graph Collaborative Filtering

In the realm of collaborative filtering recommendation systems, Graph Neural Networks (GNNs) have demonstrated remarkable performance but face significant challenges in deployment on resource-constrained edge devices due to their high embedding parameter requirements and computational costs. Using common quantization method directly on node embeddings may overlooks their graph based structure, causing error accumulation during message passing and degrading the quality of quantized embeddings.To address this, we propose Graph based Node-Aware Dynamic Quantization training for collaborative filtering (GNAQ), a novel quantization approach that leverages graph structural information to enhance the balance between efficiency and accuracy of GNNs for Top-K recommendation. GNAQ introduces a node-aware dynamic quantization strategy that adapts quantization scales to individual node embeddings by incorporating graph interaction relationships. Specifically, it initializes quantization intervals based on node-wise feature distributions and dynamically refines them through message passing in GNN layers. This approach mitigates information loss caused by fixed quantization scales and captures hierarchical semantic features in user-item interaction graphs. Additionally, GNAQ employs graph relation-aware gradient estimation to replace traditional straight-through estimators, ensuring more accurate gradient propagation during training. Extensive experiments on four real-world datasets demonstrate that GNAQ outperforms state-of-the-art quantization methods, including BiGeaR and N2UQ, by achieving average improvement in 27.8\% Recall@10 and 17.6\% NDCG@10 under 2-bit quantization. In particular, GNAQ is capable of maintaining the performance of full-precision models while reducing their model sizes by 8 to 12 times; in addition, the training time is twice as fast compared to quantization baseline methods.

cs.IR↗

arXiv2025

Compositional Zero-shot Learning via Progressive Language-based Observations

Compositional zero-shot learning aims to recognize unseen state-object compositions by leveraging known primitives (state and object) during training. However, effectively modeling interactions between primitives and generalizing knowledge to novel compositions remains a perennial challenge. There are two key factors: object-conditioned and state-conditioned variance, i.e., the appearance of states (or objects) can vary significantly when combined with different objects (or states). For instance, the state "old" can signify a vintage design for a "car" or an advanced age for a "cat". In this paper, we argue that these variances can be mitigated by predicting composition categories based on pre-observed primitive. To this end, we propose Progressive Language-based Observations (PLO), which can dynamically determine a better observation order of primitives. These observations comprise a series of concepts or languages that allow the model to understand image content in a step-by-step manner. Specifically, PLO adopts pre-trained vision-language models (VLMs) to empower the model with observation capabilities. We further devise two variants: 1) PLO-VLM: a two-step method, where a pre-observing classifier dynamically determines the observation order of two primitives. 2) PLO-LLM: a multi-step scheme, which utilizes large language models (LLMs) to craft composition-specific prompts for step-by-step observing. Extensive ablations on three challenging datasets demonstrate the superiority of PLO compared with state-of-the-art methods, affirming its abilities in compositional recognition.

cs.CV↗

arXiv2025

Generating Dialogues from Egocentric Instructional Videos for Task Assistance: Dataset, Method and Benchmark

Many everyday tasks ranging from fixing appliances, cooking recipes to car maintenance require expert knowledge, especially when tasks are complex and multi-step. Despite growing interest in AI agents, there is a scarcity of dialogue-video datasets grounded for real world task assistance. In this paper, we propose a simple yet effective approach that transforms single-person instructional videos into task-guidance two-person dialogues, aligned with fine grained steps and video-clips. Our fully automatic approach, powered by large language models, offers an efficient alternative to the substantial cost and effort required for human-assisted data collection. Using this technique, we build HowToDIV, a large-scale dataset containing 507 conversations, 6636 question-answer pairs and 24 hours of videoclips across diverse tasks in cooking, mechanics, and planting. Each session includes multi-turn conversation where an expert teaches a novice user how to perform a task step by step, while observing user's surrounding through a camera and microphone equipped wearable device. We establish the baseline benchmark performance on HowToDIV dataset through Gemma-3 model for future research on this new task of dialogues for procedural-task assistance.

cs.CV↗

arXiv2025

A Recursive Total Least Squares Solution for Bearing-Only Target Motion Analysis and Circumnavigation

Bearing-only Target Motion Analysis (TMA) is a promising technique for passive tracking in various applications as a bearing angle is easy to measure. Despite its advantages, bearing-only TMA is challenging due to the nonlinearity of the bearing measurement model and the lack of range information, which impairs observability and estimator convergence. This paper addresses these issues by proposing a Recursive Total Least Squares (RTLS) method for online target localization and tracking using mobile observers. The RTLS approach, inspired by previous results on Total Least Squares (TLS), mitigates biases in position estimation and improves computational efficiency compared to pseudo-linear Kalman filter (PLKF) methods. Additionally, we propose a circumnavigation controller to enhance system observability and estimator convergence by guiding the mobile observer in orbit around the target. Extensive simulations and experiments are performed to demonstrate the effectiveness and robustness of the proposed method. The proposed algorithm is also compared with the state-of-the-art approaches, which confirms its superior performance in terms of both accuracy and stability.

cs.RO↗

arXiv2025

Zero-shot Compositional Action Recognition with Neural Logic Constraints

Zero-shot compositional action recognition (ZS-CAR) aims to identify unseen verb-object compositions in the videos by exploiting the learned knowledge of verb and object primitives during training. Despite compositional learning's progress in ZS-CAR, two critical challenges persist: 1) Missing compositional structure constraint, leading to spurious correlations between primitives; 2) Neglecting semantic hierarchy constraint, leading to semantic ambiguity and impairing the training process. In this paper, we argue that human-like symbolic reasoning offers a principled solution to these challenges by explicitly modeling compositional and hierarchical structured abstraction. To this end, we propose a logic-driven ZS-CAR framework LogicCAR that integrates dual symbolic constraints: Explicit Compositional Logic and Hierarchical Primitive Logic. Specifically, the former models the restrictions within the compositions, enhancing the compositional reasoning ability of our model. The latter investigates the semantical dependencies among different primitives, empowering the models with fine-to-coarse reasoning capacity. By formalizing these constraints in first-order logic and embedding them into neural network architectures, LogicCAR systematically bridges the gap between symbolic abstraction and existing models. Extensive experiments on the Sth-com dataset demonstrate that our LogicCAR outperforms existing baseline methods, proving the effectiveness of our logic-driven constraints.

cs.CV↗

arXiv2025

Real-time scattering and freeze-out dynamics in Rydberg-atom lattice gauge theory

Understanding the non-equilibrium dynamics of gauge theories remains a fundamental challenge in high-energy physics. Indeed, most large scale experiments on gauge theories intrinsically rely on very far-from equilibrium dynamics, from heavy-ion to lepton and hadron collisions, which is in general extremely challenging to treat ab initio. Quantum simulation holds intriguing potential in tackling this problem and pioneering experiments have observed different characteristic features of gauge theories, such as string breaking and false vacuum decay. Here, using a programmable Rydberg atom array, we observe real-time scattering and freeze-out dynamics in a (1+1)-dimensional U(1) lattice gauge theory. Through spatiotemporal Hamiltonian engineering, we demonstrate dynamical confinement-deconfinement transitions, revealing string fragmentation and symmetry restoration during quenches. We track scattering processes with single-site resolution across a range of parameter regimes. Utilizing a double quench protocol, we observe dynamical freeze-out: upon quenching the Hamiltonian after scattering, despite the injection of an extensive energy, the system evolution -- in terms of both low-order correlations and entanglement -- freezes, effectively stabilizing a highly correlated equilibrium state -- a situation that reminisces that of collisions between heavy ions. Our work establishes a high-resolution approach for probing non-perturbative gauge dynamics, opening alternative pathways toward studying far-from-equilibrium phenomena in high-energy physics.

cond-mat.quant-gas↗

arXiv2025

Zero-Residual Concept Erasure via Progressive Alignment in Text-to-Image Model

Concept Erasure, which aims to prevent pretrained text-to-image models from generating content associated with semantic-harmful concepts (i.e., target concepts), is getting increased attention. State-of-the-art methods formulate this task as an optimization problem: they align all target concepts with semantic-harmless anchor concepts, and apply closed-form solutions to update the model accordingly. While these closed-form methods are efficient, we argue that existing methods have two overlooked limitations: 1) They often result in incomplete erasure due to "non-zero alignment residual", especially when text prompts are relatively complex. 2) They may suffer from generation quality degradation as they always concentrate parameter updates in a few deep layers. To address these issues, we propose a novel closed-form method ErasePro: it is designed for more complete concept erasure and better preserving overall generative quality. Specifically, ErasePro first introduces a strict zero-residual constraint into the optimization objective, ensuring perfect alignment between target and anchor concept features and enabling more complete erasure. Secondly, it employs a progressive, layer-wise update strategy that gradually transfers target concept features to those of the anchor concept from shallow to deep layers. As the depth increases, the required parameter changes diminish, thereby reducing deviations in sensitive deep layers and preserving generative quality. Empirical results across different concept erasure tasks (including instance, art style, and nudity erasure) have demonstrated the effectiveness of our ErasePro.

cs.CV↗

arXiv2025

Results of 15-Year Pulsar Timing of PSR J0007+7303 with Fermi-LAT

The study of pulsar glitches provides a unique window into the internal structure and dynamic processes of neutron stars. PSR J0007+7303, a very bright gamma-ray pulsar, is the first pulsar discovered by the Fermi-LAT telescope. In this paper, we present the 15 years of timing results of this pulsar using the Fermi-LAT data. We identified nine glitches, five of which are newly discovered. Among these, two are small glitches, occurring between the three previously reported ones, while the other four are large glitches. The glitches exhibit fractional frequency changes ranging from 15 x 10^-9 to 1238 x 10^-9, with intervals of approximately 1-2 years between events. Uniquely, this pulsar shows no exponential recovery behavior following any glitch, setting it apart from most glitching pulsars. Furthermore, no significant changes were observed in the gamma-ray pulse profile, flux, or phase-averaged spectra before and after glitches, indicating the stability of the pulsar's emission properties despite internal changes. A parametric analysis of the glitches yielded a fractional moment of inertia of the crustal superfluid involved in glitches as 1.06 percent, which matches extremely well with previous statistical work if the non-dissipative entrainment effect is not considered and strongly supports the internal origin of these glitches. These results highlight the distinct glitch behavior of PSR J0007+7303 and offer valuable insights into the crust-superfluid interaction in neutron stars. The physical origin of no exponential recovery is also discussed.

astro-ph.HE↗

arXiv2025

Enhancing Transferability and Consistency in Cross-Domain Recommendations via Supervised Disentanglement

Cross-domain recommendation (CDR) aims to alleviate the data sparsity by transferring knowledge across domains. Disentangled representation learning provides an effective solution to model complex user preferences by separating intra-domain features (domain-shared and domain-specific features), thereby enhancing robustness and interpretability. However, disentanglement-based CDR methods employing generative modeling or GNNs with contrastive objectives face two key challenges: (i) pre-separation strategies decouple features before extracting collaborative signals, disrupting intra-domain interactions and introducing noise; (ii) unsupervised disentanglement objectives lack explicit task-specific guidance, resulting in limited consistency and suboptimal alignment. To address these challenges, we propose DGCDR, a GNN-enhanced encoder-decoder framework. To handle challenge (i), DGCDR first applies GNN to extract high-order collaborative signals, providing enriched representations as a robust foundation for disentanglement. The encoder then dynamically disentangles features into domain-shared and -specific spaces, preserving collaborative information during the separation process. To handle challenge (ii), the decoder introduces an anchor-based supervision that leverages hierarchical feature relationships to enhance intra-domain consistency and cross-domain alignment. Extensive experiments on real-world datasets demonstrate that DGCDR achieves state-of-the-art performance, with improvements of up to 11.59% across key metrics. Qualitative analyses further validate its superior disentanglement quality and transferability. Our source code and datasets are available on GitHub for further comparison.

cs.IR↗

arXiv2025

Cooperative Circumnavigation for Multi-Quadrotor Systems via Onboard Sensing

A cooperative circumnavigation framework is proposed for multi-quadrotor systems to enclose and track a moving target without reliance on external localization systems. The distinct relationships between quadrotor-quadrotor and quadrotor-target interactions are evaluated using a heterogeneous perception strategy and corresponding state estimation algorithms. A modified Kalman filter is developed to fuse visual-inertial odometry with range measurements to enhance the accuracy of inter-quadrotor relative localization. An event-triggered distributed Kalman filter is designed to achieve robust target state estimation under visual occlusion by incorporating neighbor measurements and estimated inter-quadrotor relative positions. Using the estimation results, a cooperative circumnavigation controller is constructed, leveraging an oscillator-based autonomous formation flight strategy. We conduct extensive indoor and outdoor experiments to validate the efficiency of the proposed circumnavigation framework in occluded environments. Furthermore, a quadrotor failure experiment highlights the inherent fault tolerance property of the proposed framework, underscoring its potential for deployment in search-and-rescue operations.

cs.RO↗

arXiv2025

Pseudo Labels-based Neural Speech Enhancement for the AVSR Task in the MISP-Meeting Challenge

This paper presents our system for the MISP-Meeting Challenge Track 2. The primary difficulty lies in the dataset, which contains strong background noise, reverberation, overlapping speech, and diverse meeting topics. To address these issues, we (a) designed G-SpatialNet, a speech enhancement (SE) model to improve Guided Source Separation (GSS) signals; (b) proposed TLS, a framework comprising time alignment, level alignment, and signal-to-noise ratio filtering, to generate signal-level pseudo labels for real-recorded far-field audio data, thereby facilitating SE models' training; and (c) explored fine-tuning strategies, data augmentation, and multimodal information to enhance the performance of pre-trained Automatic Speech Recognition (ASR) models in meeting scenarios. Finally, our system achieved character error rates (CERs) of 5.44% and 9.52% on the Dev and Eval sets, respectively, with relative improvements of 64.8% and 52.6% over the baseline, securing second place.

cs.SD↗