arXiv subjects

Chang Liu

Publications and source records attributed to Chang Liu.

At least 649 records · Page 36Linked to original sources

arXiv2024

AnatoMix: Anatomy-aware Data Augmentation for Multi-organ Segmentation

Multi-organ segmentation in medical images is a widely researched task and can save much manual efforts of clinicians in daily routines. Automating the organ segmentation process using deep learning (DL) is a promising solution and state-of-the-art segmentation models are achieving promising accuracy. In this work, We proposed a novel data augmentation strategy for increasing the generalizibility of multi-organ segmentation datasets, namely AnatoMix. By object-level matching and manipulation, our method is able to generate new images with correct anatomy, i.e. organ segmentation mask, exponentially increasing the size of the segmentation dataset. Initial experiments have been done to investigate the segmentation performance influenced by our method on a public CT dataset. Our augmentation method can lead to mean dice of 76.1, compared with 74.8 of the baseline method.

eess.IV↗

arXiv2024

Intelligent Grimm -- Open-ended Visual Storytelling via Latent Diffusion Models

Generative models have recently exhibited exceptional capabilities in text-to-image generation, but still struggle to generate image sequences coherently. In this work, we focus on a novel, yet challenging task of generating a coherent image sequence based on a given storyline, denoted as open-ended visual storytelling. We make the following three contributions: (i) to fulfill the task of visual storytelling, we propose a learning-based auto-regressive image generation model, termed as StoryGen, with a novel vision-language context module, that enables to generate the current frame by conditioning on the corresponding text prompt and preceding image-caption pairs; (ii) to address the data shortage of visual storytelling, we collect paired image-text sequences by sourcing from online videos and open-source E-books, establishing processing pipeline for constructing a large-scale dataset with diverse characters, storylines, and artistic styles, named StorySalon; (iii) Quantitative experiments and human evaluations have validated the superiority of our StoryGen, where we show StoryGen can generalize to unseen characters without any optimization, and generate image sequences with coherent content and consistent character. Code, dataset, and models are available at https://haoningwu3639.github.io/StoryGen_Webpage/

cs.CV↗

arXiv2024

ASPIRe: An Informative Trajectory Planner with Mutual Information Approximation for Target Search and Tracking

This paper proposes an informative trajectory planning approach, namely, \textit{adaptive particle filter tree with sigma point-based mutual information reward approximation} (ASPIRe), for mobile target search and tracking (SAT) in cluttered environments with limited sensing field of view. We develop a novel sigma point-based approximation to accurately estimate mutual information (MI) for general, non-Gaussian distributions utilizing particle representation of the belief state, while simultaneously maintaining high computational efficiency. Building upon the MI approximation, we develop the Adaptive Particle Filter Tree (APFT) approach with MI as the reward, which features belief state tree nodes for informative trajectory planning in continuous state and measurement spaces. An adaptive criterion is proposed in APFT to adjust the planning horizon based on the expected information gain. Simulations and physical experiments demonstrate that ASPIRe achieves real-time computation and outperforms benchmark methods in terms of both search efficiency and estimation accuracy.

cs.RO↗

arXiv2024

Random Green's function method for large-scale electronic structure calculation

We report a linear-scaling random Green's function (rGF) method for large-scale electronic structure calculation. In this method, the rGF is defined on a set of random states to stochastically express the density matrix, and rGF is calculated with the linear-scaling computational cost. We show the rGF method is generally applicable to the nonorthogonal localized basis, and circumvent the large Chebyshev expansion for the density matrix. As a demonstration, we implement rGF with density-functional Tight-Binding method and apply it to self-consistently calculate water clusters up 9984 H2Os. We find the rGF method combining with a simple fragment correction can reach an error of ~1meV per H2O in total energy, compared to the deterministic calculations, due to the self-average. The development of rGF method advances the stochastic electronic structure theory to a new stage of the efficiency and applicability.

cond-mat.mes-hall↗

arXiv2024

Path Planning based on 2D Object Bounding-box

The implementation of Autonomous Driving (AD) technologies within urban environments presents significant challenges. These challenges necessitate the development of advanced perception systems and motion planning algorithms capable of managing situations of considerable complexity. Although the end-to-end AD method utilizing LiDAR sensors has achieved significant success in this scenario, we argue that its drawbacks may hinder its practical application. Instead, we propose the vision-centric AD as a promising alternative offering a streamlined model without compromising performance. In this study, we present a path planning method that utilizes 2D bounding boxes of objects, developed through imitation learning in urban driving scenarios. This is achieved by integrating high-definition (HD) map data with images captured by surrounding cameras. Subsequent perception tasks involve bounding-box detection and tracking, while the planning phase employs both local embeddings via Graph Neural Network (GNN) and global embeddings via Transformer for temporal-spatial feature aggregation, ultimately producing optimal path planning information. We evaluated our model on the nuPlan planning task and observed that it performs competitively in comparison to existing vision-centric methods.

cs.RO↗

arXiv2024

RECALL+: Adversarial Web-based Replay for Continual Learning in Semantic Segmentation

Catastrophic forgetting of previous knowledge is a critical issue in continual learning typically handled through various regularization strategies. However, existing methods struggle especially when several incremental steps are performed. In this paper, we extend our previous approach (RECALL) and tackle forgetting by exploiting unsupervised web-crawled data to retrieve examples of old classes from online databases. In contrast to the original methodology, which did not incorporate an assessment of web-based data, the present work proposes two advanced techniques: an adversarial approach and an adaptive threshold strategy. These methods are utilized to meticulously choose samples from web data that exhibit strong statistical congruence with the no longer available training data. Furthermore, we improved the pseudo-labeling scheme to achieve a more accurate labeling of web data that also considers classes being learned in the current step. Experimental results show that this enhanced approach achieves remarkable results, particularly when the incremental scenario spans multiple steps.

cs.CV↗

arXiv2024

A Novel Multivariate Skew-Normal Mixture Model and Its Application in Path-Planning for Very-Large-Scale Robotic Systems

This paper addresses the path-planning challenge for very large-scale robotic systems (VLSR) operating in complex and cluttered environments. VLSR systems consist of numerous cooperative agents or robots working together autonomously. Traditionally, many approaches for VLSR systems are developed based on Gaussian mixture models (GMMs), where the GMMs represent agents' evolving spatial distribution, serving as a macroscopic view of the system's state. However, our recent research into VLSR systems has unveiled limitations in using GMMs to represent agent distributions, especially in cluttered environments. To overcome these limitations, we propose a novel model called the skew-normal mixture model (SNMM) for representing agent distributions. Additionally, we present a parameter learning algorithm designed to estimate the SNMM's parameters using sample data. Furthermore, we develop two SNMM-based path-planning algorithms to guide VLSR systems through complex and cluttered environments. Our simulation results demonstrate the effectiveness and superiority of these algorithms compared to GMM-based path-planning methods.

cs.MA↗

arXiv2024

Robust Contracts with Exploration

We study a two-period moral hazard problem; there are two agents, with action sets that are unknown to the principal. The principal contracts with each agent sequentially, and seeks to maximize the worst-case discounted sum of payoffs, where the worst case is over the possible action sets. The principal observes the action chosen by the first agent, and then offers a new contract to the second agent based on this knowledge, thus having the opportunity to explore in the first period. We introduce and compare three different notions of dynamic worst-case considerations. Within each notion, we define a suitable rule of updating and characterize the principal's optimal payoff guarantee. We find that linear contracts are robustly optimal not only in static settings, but also in dynamic environments with exploration.

econ.TH↗

arXiv2024

Chemical inhomogeneities in high-entropy alloys help mitigate the strength-ductility trade-off

Metallurgists have long been accustomed to a trade-off between yield strength and tensile ductility. Extending previously known strain-hardening mechanisms, the emerging multi-principal-element alloys (MPEAs) offer additional help in promoting the strength-ductility synergy, towards gigapascal yield strength simultaneously with pure-metal-like tensile ductility. The highly concentrated chemical make-up in these 'high-entropy' alloys (HEAs) adds, at ultrafine spatial scale from sub-nanometer to tens of nanometers, inherent chemical inhomogeneities in local composition and local chemical order (LCO). These institute a 'nano-cocktail' environment that exerts extra dragging forces, rendering a much wavier motion of dislocation lines (in stick-slip mode) different from dilute solutions. The variable fault energy landscape also makes the dislocation movement sluggish, increasing their chances to hit one another and react to increase entanglement. The accumulation of dislocations (plus faults) dynamically stores obstacles against ensuing dislocation motion to sustain an adequate strain-hardening rate at high flow stresses, delaying plastic instability to enable large (uniform) elongation. The successes summarized advocate MPEAs as an effective recipe towards ultrahigh strength at little expense of tensile ductility. The insight gained also answers the question as to what new mechanical behavior the HEAs have to offer, beyond what has been well documented for traditional metals and solid solutions.

cond-mat.mtrl-sci↗

arXiv2024

Transformer-empowered Multi-modal Item Embedding for Enhanced Image Search in E-Commerce

Over the past decade, significant advances have been made in the field of image search for e-commerce applications. Traditional image-to-image retrieval models, which focus solely on image details such as texture, tend to overlook useful semantic information contained within the images. As a result, the retrieved products might possess similar image details, but fail to fulfil the user's search goals. Moreover, the use of image-to-image retrieval models for products containing multiple images results in significant online product feature storage overhead and complex mapping implementations. In this paper, we report the design and deployment of the proposed Multi-modal Item Embedding Model (MIEM) to address these limitations. It is capable of utilizing both textual information and multiple images about a product to construct meaningful product features. By leveraging semantic information from images, MIEM effectively supplements the image search process, improving the overall accuracy of retrieval results. MIEM has become an integral part of the Shopee image search platform. Since its deployment in March 2023, it has achieved a remarkable 9.90% increase in terms of clicks per user and a 4.23% boost in terms of orders per user for the image search feature on the Shopee e-commerce platform.

cs.CV↗

arXiv2024

VoroNav: Voronoi-based Zero-shot Object Navigation with Large Language Model

In the realm of household robotics, the Zero-Shot Object Navigation (ZSON) task empowers agents to adeptly traverse unfamiliar environments and locate objects from novel categories without prior explicit training. This paper introduces VoroNav, a novel semantic exploration framework that proposes the Reduced Voronoi Graph to extract exploratory paths and planning nodes from a semantic map constructed in real time. By harnessing topological and semantic information, VoroNav designs text-based descriptions of paths and images that are readily interpretable by a large language model (LLM). In particular, our approach presents a synergy of path and farsight descriptions to represent the environmental context, enabling LLM to apply commonsense reasoning to ascertain waypoints for navigation. Extensive evaluation on HM3D and HSSD validates VoroNav surpasses existing benchmarks in both success rate and exploration efficiency (absolute improvement: +2.8% Success and +3.7% SPL on HM3D, +2.6% Success and +3.8% SPL on HSSD). Additionally introduced metrics that evaluate obstacle avoidance proficiency and perceptual efficiency further corroborate the enhancements achieved by our method in ZSON planning. Project page: https://voro-nav.github.io

cs.RO↗

arXiv2024

Prospects for detecting neutron star-white dwarf mergers with decihertz gravitational-wave observatories

Based on different neutron star-white dwarf (NS-WD) population models, we investigate the prospects of gravitational-wave (GW) detections for NS-WD mergers, with the help of early warnings from two space-borne decihertz GW observatories, DO-Optimal and DECIGO. We not only give quick assessments of the GW detection rates for NS-WD mergers with the two decihertz GW detectors, but also report systematic analyses on the characteristics of GW-detectable merger events using the method of Fisher matrix. With a sufficient one-day early-warning time, the yearly GW detection number for DO-Optimal is in the range of $ (1.5$-$1.9) \times 10^{3}$, while it is $ (3.3$-$4.6) \times 10^{4}$ for DECIGO. More importantly, our results show that most NS-WD mergers can be localized with an uncertainty of $O(10^{-2})\,\mathrm{deg}^2$. Given the NS-WD merger as a possible origin for a peculiar long-duration gamma-ray burst, GRB 211211A, followed with kilonova-like emissions, we further suggest that the GW early-warning detection would allow future electromagnetic telescopes to get prepared to follow-up transients after some special NS-WD mergers. Based on our analyses, we emphasize that such a feasible "wait-for" pattern can help to firmly identify the origin of GRB 211211A-like events in the future and bring excellent opportunities for the multimessenger astronomy.

astro-ph.HE↗

arXiv2024

Two-View Topogram-Based Anatomy-Guided CT Reconstruction for Prospective Risk Minimization

To facilitate a prospective estimation of CT effective dose and risk minimization process, a prospective spatial dose estimation and the known anatomical structures are expected. To this end, a CT reconstruction method is required to reconstruct CT volumes from as few projections as possible, i.e. by using the topograms, with anatomical structures as correct as possible. In this work, an optimized CT reconstruction model based on a generative adversarial network (GAN) is proposed. The GAN is trained to reconstruct 3D volumes from an anterior-posterior and a lateral CT projection. To enhance anatomical structures, a pre-trained organ segmentation network and the 3D perceptual loss are applied during the training phase, so that the model can then generate both organ-enhanced CT volume and the organ segmentation mask. The proposed method can reconstruct CT volumes with PSNR of 26.49, RMSE of 196.17, and SSIM of 0.64, compared to 26.21, 201.55 and 0.63 using the baseline method. In terms of the anatomical structure, the proposed method effectively enhances the organ shape and boundary and allows for a straight-forward identification of the relevant anatomical structures. We note that conventional reconstruction metrics fail to indicate the enhancement of anatomical structures. In addition to such metrics, the evaluation is expanded with assessing the organ segmentation performance. The average organ dice of the proposed method is 0.71 compared with 0.63 in baseline model, indicating the enhancement of anatomical structures.

eess.IV↗

arXiv2024

PepHarmony: A Multi-View Contrastive Learning Framework for Integrated Sequence and Structure-Based Peptide Encoding

Recent advances in protein language models have catalyzed significant progress in peptide sequence representation. Despite extensive exploration in this field, pre-trained models tailored for peptide-specific needs remain largely unaddressed due to the difficulty in capturing the complex and sometimes unstable structures of peptides. This study introduces a novel multi-view contrastive learning framework PepHarmony for the sequence-based peptide encoding task. PepHarmony innovatively combines both sequence- and structure-level information into a sequence-level encoding module through contrastive learning. We carefully select datasets from the Protein Data Bank (PDB) and AlphaFold database to encompass a broad spectrum of peptide sequences and structures. The experimental data highlights PepHarmony's exceptional capability in capturing the intricate relationship between peptide sequences and structures compared with the baseline and fine-tuned models. The robustness of our model is confirmed through extensive ablation studies, which emphasize the crucial roles of contrastive loss and strategic data sorting in enhancing predictive performance. The proposed PepHarmony framework serves as a notable contribution to peptide representations, and offers valuable insights for future applications in peptide drug discovery and peptide engineering. We have made all the source code utilized in this study publicly accessible via GitHub at https://github.com/zhangruochi/PepHarmony or http://www.healthinformaticslab.org/supp/.

cs.LG↗

arXiv2024

Error bounds of constant gain least-mean-squares algorithms

Constant gain least-mean-squares (LMS) algorithms have a wide range of applications in trajectory tracking problems, but the formal convergence of LMS in mean square is not yet fully established. This work provides an upper bound on the constant gain that guarantees a bounded mean-squared error of LMS for a general design vector. These results highlight the role of the fourth-order moment of the design vector. Numerical examples demonstrate the applicability of this upper bound in setting a constant gain in LMS, while existing criteria may fail. We also provide the associated error bound, which can be applied to design vectors with linearly dependent elements.

eess.SP↗

arXiv2024

Instance Brownian Bridge as Texts for Open-vocabulary Video Instance Segmentation

Temporally locating objects with arbitrary class texts is the primary pursuit of open-vocabulary Video Instance Segmentation (VIS). Because of the insufficient vocabulary of video data, previous methods leverage image-text pretraining model for recognizing object instances by separately aligning each frame and class texts, ignoring the correlation between frames. As a result, the separation breaks the instance movement context of videos, causing inferior alignment between video and text. To tackle this issue, we propose to link frame-level instance representations as a Brownian Bridge to model instance dynamics and align bridge-level instance representation to class texts for more precisely open-vocabulary VIS (BriVIS). Specifically, we build our system upon a frozen video segmentor to generate frame-level instance queries, and design Temporal Instance Resampler (TIR) to generate queries with temporal context from frame queries. To mold instance queries to follow Brownian bridge and accomplish alignment with class texts, we design Bridge-Text Alignment (BTA) to learn discriminative bridge-level representations of instances via contrastive objectives. Setting MinVIS as the basic video segmentor, BriVIS surpasses the Open-vocabulary SOTA (OV2Seg) by a clear margin. For example, on the challenging large-vocabulary VIS dataset (BURST), BriVIS achieves 7.43 mAP and exhibits 49.49% improvement compared to OV2Seg (4.97 mAP).

cs.CV↗

arXiv2024

Anatomy of parameter-estimation biases in overlapping gravitational-wave signals

In future gravitational-wave (GW) detections, a large number of overlapping GW signals will appear in the data stream of detectors. When extracting information from one signal, the presence of other signals can cause large parameter estimation biases. Using the Fisher matrix (FM), we develop a bias analysis procedure to investigate how each parameter of other signals affects the inference biases. Taking two-signal overlapping as an example, we show detailedly and quantitatively that the biases essentially originate from the overlapping of the frequency evolution. Furthermore, we find that the behaviors of the correlation coefficients between the parameters of the two signals are similar to the biases. Both of them can be used as characterization of the influence between signals. We also corroborate the bias results of the FM method with full Bayesian analysis. Our results can provide guidance for the development of new PE algorithms on overlapping signals, and the analysis methodology has the potential to generalize.

astro-ph.IM↗

arXiv2024

Shell-infill composite structure design based on a hybrid explicit-implicit topology optimization method

The present paper introduces a hybrid explicit-implicit topology optimization method for shell-infill composite structure design. The proposed approach effectively combines the advantages of the explicit Moving Morphable Component (MMC) method, which describes structural topology only using a set of geometric parameters, and the implicit Solid Isotropic Material with Penalization (SIMP) method, which offers greater design freedom for characterizing the structural features. Compared to the existing methods for shell-infill structure design, the proposed approach can obtain optimized shell-infill structures with complex infill topology without resorting to complex filtering/projection operations. Numerical examples demonstrate the effectiveness of the proposed approach.

math.OC↗