arXiv subjects

Ming Cheng

Publications and source records attributed to Ming Cheng.

At least 55 records · Page 3Linked to original sources

arXiv2025

Expanding solutions near unstable Lane-Emden stars

We consider the compressible Euler-Poisson equations for polytropes $P(ρ)=Kρ^γ$ with $γ\in \left(\frac{6}{5},\frac{4}{3} \right]$ and the white dwarf stars. For $γ=\frac{4}{3},$ we establish the existence of a global weak solution for the spherically symmetric initial data with mass less than the mass of the Lane-Emden stars (i.e. non-rotating polytropes). For $γ\in \left(\frac{6}{5},\frac{4}{3} \right)$, we show the existence of global weak solution for spherical symmetric initial data in an invariant set containing a neighborhood of Lane-Emden stars. Moreover, the support of these solution expands to infinity. As a corollary, this proves the strong instability of the Lane-Emden stars for $γ\in \left( \frac{6}{5},\frac{4}{3}\right] $. For $γ\in \left(\frac{6}{5},\frac{4}{3} \right),$ our results provide the first example of expanding solutions near the Lane-Emden stars. For white dwarf stars, we prove that the solution cannot collapse if the mass of initial data is less than the Chandrasekhar limit mass, which is the supremum of the mass of the non-rotating white dwarf stars. Our proof strongly uses the variational characterization of the Lane-Emden stars. First, we relate the best constant of a Hardy-Littlewood type inequality with the mass of the Lane-Emden stars with $γ=\frac{4}{3}$, which is further shown to equal the Chandrasekhar limit mass. For $γ\in\left( \frac{6}{5},\frac{4}{3}\right) $, we show that the Lane-Emden stars are minimizers of an energy-mass functional subject to a Pohozaev type constraint. This is crucial in the construction of the invariant set of expanding solutions.

math.AP↗

arXiv2025

GS-CPR: Efficient Camera Pose Refinement via 3D Gaussian Splatting

We leverage 3D Gaussian Splatting (3DGS) as a scene representation and propose a novel test-time camera pose refinement (CPR) framework, GS-CPR. This framework enhances the localization accuracy of state-of-the-art absolute pose regression and scene coordinate regression methods. The 3DGS model renders high-quality synthetic images and depth maps to facilitate the establishment of 2D-3D correspondences. GS-CPR obviates the need for training feature extractors or descriptors by operating directly on RGB images, utilizing the 3D foundation model, MASt3R, for precise 2D matching. To improve the robustness of our model in challenging outdoor environments, we incorporate an exposure-adaptive module within the 3DGS framework. Consequently, GS-CPR enables efficient one-shot pose refinement given a single RGB query and a coarse initial pose estimation. Our proposed approach surpasses leading NeRF-based optimization methods in both accuracy and runtime across indoor and outdoor visual localization benchmarks, achieving new state-of-the-art accuracy on two indoor datasets. The project page is available at https://xrim-lab.github.io/GS-CPR/.

cs.CV↗

arXiv2025

VTechAGP: An Academic-to-General-Audience Text Paraphrase Dataset and Benchmark Models

Existing text simplification or paraphrase datasets mainly focus on sentence-level text generation in a general domain. These datasets are typically developed without using domain knowledge. In this paper, we release a novel dataset, VTechAGP, which is the first academic-to-general-audience text paraphrase dataset consisting of document-level these and dissertation academic and general-audience abstract pairs from 8 colleges authored over 25 years. We also propose a novel dynamic soft prompt generative language model, DSPT5. For training, we leverage a contrastive-generative loss function to learn the keyword vectors in the dynamic prompt. For inference, we adopt a crowd-sampling decoding strategy at both semantic and structural levels to further select the best output candidate. We evaluate DSPT5 and various state-of-the-art large language models (LLMs) from multiple perspectives. Results demonstrate that the SOTA LLMs do not provide satisfactory outcomes, while the lightweight DSPT5 can achieve competitive results. To the best of our knowledge, we are the first to build a benchmark dataset and solutions for academic-to-general-audience text paraphrase dataset. Models will be public after acceptance.

cs.CL↗

arXiv2025

Visual Zero-Shot E-Commerce Product Attribute Value Extraction

Existing zero-shot product attribute value (aspect) extraction approaches in e-Commerce industry rely on uni-modal or multi-modal models, where the sellers are asked to provide detailed textual inputs (product descriptions) for the products. However, manually providing (typing) the product descriptions is time-consuming and frustrating for the sellers. Thus, we propose a cross-modal zero-shot attribute value generation framework (ViOC-AG) based on CLIP, which only requires product images as the inputs. ViOC-AG follows a text-only training process, where a task-customized text decoder is trained with the frozen CLIP text encoder to alleviate the modality gap and task disconnection. During the zero-shot inference, product aspects are generated by the frozen CLIP image encoder connected with the trained task-customized text decoder. OCR tokens and outputs from a frozen prompt-based LLM correct the decoded outputs for out-of-domain attribute values. Experiments show that ViOC-AG significantly outperforms other fine-tuned vision-language models for zero-shot attribute value extraction.

cs.IR↗

arXiv2025

Learning Musical Representations for Music Performance Question Answering

Music performances are representative scenarios for audio-visual modeling. Unlike common scenarios with sparse audio, music performances continuously involve dense audio signals throughout. While existing multimodal learning methods on the audio-video QA demonstrate impressive capabilities in general scenarios, they are incapable of dealing with fundamental problems within the music performances: they underexplore the interaction between the multimodal signals in performance and fail to consider the distinctive characteristics of instruments and music. Therefore, existing methods tend to answer questions regarding musical performances inaccurately. To bridge the above research gaps, (i) given the intricate multimodal interconnectivity inherent to music data, our primary backbone is designed to incorporate multimodal interactions within the context of music; (ii) to enable the model to learn music characteristics, we annotate and release rhythmic and music sources in the current music datasets; (iii) for time-aware audio-visual modeling, we align the model's music predictions with the temporal dimension. Our experiments show state-of-the-art effects on the Music AVQA datasets. Our code is available at https://github.com/xid32/Amuse.

cs.CV↗

arXiv2025

Temporal Working Memory: Query-Guided Segment Refinement for Enhanced Multimodal Understanding

Multimodal foundation models (MFMs) have demonstrated significant success in tasks such as visual captioning, question answering, and image-text retrieval. However, these models face inherent limitations due to their finite internal capacity, which restricts their ability to process extended temporal sequences, a crucial requirement for comprehensive video and audio analysis. To overcome these challenges, we introduce a specialized cognitive module, temporal working memory (TWM), which aims to enhance the temporal modeling capabilities of MFMs. It selectively retains task-relevant information across temporal dimensions, ensuring that critical details are preserved throughout the processing of video and audio content. The TWM uses a query-guided attention approach to focus on the most informative multimodal segments within temporal sequences. By retaining only the most relevant content, TWM optimizes the use of the model's limited capacity, enhancing its temporal modeling ability. This plug-and-play module can be easily integrated into existing MFMs. With our TWM, nine state-of-the-art models exhibit significant performance improvements across tasks such as video captioning, question answering, and video-text retrieval. By enhancing temporal modeling, TWM extends the capability of MFMs to handle complex, time-sensitive data effectively. Our code is available at https://github.com/xid32/NAACL_2025_TWM.

cs.CV↗

arXiv2025

Turning point principle for stability of viscous gaseous stars

We consider stability of non-rotating viscous gaseous stars modeled by the Navier-Stokes-Poisson system. Under general assumptions on the equations of states, we proved that the number of unstable modes for the linearized Navier-Stokes-Poisson system equals that of the linearized Euler-Poisson system modeling inviscid gaseous stars. In particular, the turning point principle holds true for non-rotating stars with or without viscosity. That is, the transition of stability only occurs at the extrema of the total mass and the number of unstable modes is determined by the mass-radius curve. For the proof, we establish an infinite dimensional Kelvin-Tait-Chetaev theorem for a class of linear second order PDEs with dissipation. Moreover, we prove that linear stability implies nonlinear asymptotic stability and linear instability implies nonlinear instability for Navier-Stokes-Poisson system.

math.AP↗

arXiv2024

FT2TF: First-Person Statement Text-To-Talking Face Generation

Talking face generation has gained immense popularity in the computer vision community, with various applications including AR, VR, teleconferencing, digital assistants, and avatars. Traditional methods are mainly audio-driven, which have to deal with the inevitable resource-intensive nature of audio storage and processing. To address such a challenge, we propose FT2TF - First-Person Statement Text-To-Talking Face Generation, a novel one-stage end-to-end pipeline for talking face generation driven by first-person statement text. Different from previous work, our model only leverages visual and textual information without any other sources (e.g., audio/landmark/pose) during inference. Extensive experiments are conducted on LRS2 and LRS3 datasets, and results on multi-dimensional evaluation metrics are reported. Both quantitative and qualitative results showcase that FT2TF outperforms existing relevant methods and reaches the state-of-the-art. This achievement highlights our model's capability to bridge first-person statements and dynamic face generation, providing insightful guidance for future work.

cs.CV↗

arXiv2024

STARFlow: Spatial Temporal Feature Re-embedding with Attentive Learning for Real-world Scene Flow

Scene flow prediction is a crucial underlying task in understanding dynamic scenes as it offers fundamental motion information. However, contemporary scene flow methods encounter three major challenges. Firstly, flow estimation solely based on local receptive fields lacks long-dependency matching of point pairs. To address this issue, we propose global attentive flow embedding to match all-to-all point pairs in both feature space and Euclidean space, providing global initialization before local refinement. Secondly, there are deformations existing in non-rigid objects after warping, which leads to variations in the spatiotemporal relation between the consecutive frames. For a more precise estimation of residual flow, a spatial temporal feature re-embedding module is devised to acquire the sequence features after deformation. Furthermore, previous methods perform poor generalization due to the significant domain gap between the synthesized and LiDAR-scanned datasets. We leverage novel domain adaptive losses to effectively bridge the gap of motion inference from synthetic to real-world. Experiments demonstrate that our approach achieves state-of-the-art performance across various datasets, with particularly outstanding results on real-world LiDAR-scanned datasets. Our code is available at https://github.com/O-VIGIA/StarFlow.

cs.CV↗

arXiv2024

Transformer for seismic image super-resolution

Seismic images obtained by stacking or migration are usually characterized as low signal-to-noise ratio (SNR), low dominant frequency and sparse sampling both in depth (or time) and offset dimensions. For improving the resolution of seismic images, we proposed a deep learning-based method to achieve super-resolution (SR) in only one step, which means performing the denoising, interpolation and frequency extrapolation at the same time. We design a seismic image super-resolution Transformer (SIST) to extract and fuse local and global features, which focuses more on the energy and extension shapes of effective events (horizons, folds and faults, etc.) from noisy seismic images. We extract the edge images of input images by Canny algorithm as masks to generate the input data with double channels, which improves the amplitude preservation and reduces the interference of noises. The residual groups containing Swin-Transformer blocks and residual connections consist of the backbone of SIST, which extract the global features in a window with preset size and decrease computational cost meanwhile. The pixel shuffle layers are used to up-sample the output feature maps from the backbone to improve the edges, meanwhile up-sampling the input data through a skip connection to enhance the amplitude preservation of the final images especially for clarifying weak events. 3-dimensional synthetic seismic volumes with complex geological structures are created, and the amplitudes of half of the volumes are mixtures of strong and weak, then select 2-dimensional slices randomly to generate training datasets which fits field data well to perform supervised learning. Both numerical tests on synthetic and field data in different exploration regions demonstrate the feasibility of our method.

physics.geo-ph↗

arXiv2024

SSRFlow: Semantic-aware Fusion with Spatial Temporal Re-embedding for Real-world Scene Flow

Scene flow, which provides the 3D motion field of the first frame from two consecutive point clouds, is vital for dynamic scene perception. However, contemporary scene flow methods face three major challenges. Firstly, they lack global flow embedding or only consider the context of individual point clouds before embedding, leading to embedded points struggling to perceive the consistent semantic relationship of another frame. To address this issue, we propose a novel approach called Dual Cross Attentive (DCA) for the latent fusion and alignment between two frames based on semantic contexts. This is then integrated into Global Fusion Flow Embedding (GF) to initialize flow embedding based on global correlations in both contextual and Euclidean spaces. Secondly, deformations exist in non-rigid objects after the warping layer, which distorts the spatiotemporal relation between the consecutive frames. For a more precise estimation of residual flow at next-level, the Spatial Temporal Re-embedding (STR) module is devised to update the point sequence features at current-level. Lastly, poor generalization is often observed due to the significant domain gap between synthetic and LiDAR-scanned datasets. We leverage novel domain adaptive losses to effectively bridge the gap of motion inference from synthetic to real-world. Experiments demonstrate that our approach achieves state-of-the-art (SOTA) performance across various datasets, with particularly outstanding results in real-world LiDAR-scanned situations. Our code will be released upon publication.

cs.CV↗

arXiv2024

Distortion Recovery: A Two-Stage Method for Guitar Effect Removal

Removing audio effects from electric guitar recordings makes it easier for post-production and sound editing. An audio distortion recovery model not only improves the clarity of the guitar sounds but also opens up new opportunities for creative adjustments in mixing and mastering. While progress have been made in creating such models, previous efforts have largely focused on synthetic distortions that may be too simplistic to accurately capture the complexities seen in real-world recordings. In this paper, we tackle the task by using a dataset of guitar recordings rendered with commercial-grade audio effect VST plugins. Moreover, we introduce a novel two-stage methodology for audio distortion recovery. The idea is to firstly process the audio signal in the Mel-spectrogram domain in the first stage, and then use a neural vocoder to generate the pristine original guitar sound from the processed Mel-spectrogram in the second stage. We report a set of experiments demonstrating the effectiveness of our approach over existing methods, through both subjective and objective evaluation metrics.

cs.SD↗

arXiv2024

VoxBlink2: A 100K+ Speaker Recognition Corpus and the Open-Set Speaker-Identification Benchmark

In this paper, we provide a large audio-visual speaker recognition dataset, VoxBlink2, which includes approximately 10M utterances with videos from 110K+ speakers in the wild. This dataset represents a significant expansion over the VoxBlink dataset, encompassing a broader diversity of speakers and scenarios by the grace of an optimized data collection pipeline. Afterward, we explore the impact of training strategies, data scale, and model complexity on speaker verification and finally establish a new single-model state-of-the-art EER at 0.170% and minDCF at 0.006% on the VoxCeleb1-O test set. Such remarkable results motivate us to explore speaker recognition from a new challenging perspective. We raise the Open-Set Speaker-Identification task, which is designed to either match a probe utterance with a known gallery speaker or categorize it as an unknown query. Associated with this task, we design concrete benchmark and evaluation protocols. The data and model resources can be found in http://voxblink2.github.io.

eess.AS↗

arXiv2024

Seismic Interpolation Transformer for Consecutively Missing Data: A Case Study in DAS-VSP Data

Distributed optical fiber acoustic sensing (DAS) is a rapidly-developed seismic acquisition technology with advantages of low cost, high resolution, high sensitivity, and small interval, etc. Nonetheless, consecutively missing cases often appear in real seismic data acquired by DAS system due to some factors, including optical fiber damage and inferior coupling between cable and well. Recently, some deep-learning seismic interpolation methods based on convolutional neural network (CNN) have shown impressive performance in regular and random missing cases but still remain the consecutively missing case as a challenging task. The main reason is that the weight sharing makes it difficult for CNN to capture enough comprehensive features. In this paper, we propose a transformer-based interpolation method, called seismic interpolation transformer (SIT), to deal with the consecutively missing case. This proposed SIT is an encoder-decoder structure connected by some U-shaped swin-transformer blocks. In encoder and decoder part, the multi-head self-attention (MSA) mechanism is used to capture global features which is essential for the reconstruction of consecutively missing traces. The U-shaped swin-transformer blocks are utilized to perform feature extraction operations on feature maps with different resolutions. Moreover, we combine the loss based on structural similarity index (SSIM) and L1 norm to propose a novel loss function for SIT. In experiments, this proposed SIT outperforms U-Net and swin-transformer. Moreover, ablation studies also demonstrate the advantages of new network architecture and loss function.

physics.geo-ph↗

arXiv2024

GluMarker: A Novel Predictive Modeling of Glycemic Control Through Digital Biomarkers

The escalating prevalence of diabetes globally underscores the need for diabetes management. Recent research highlights the growing focus on digital biomarkers in diabetes management, with innovations in computational frameworks and noninvasive monitoring techniques using personalized glucose metrics. However, they predominantly focus on insulin dosing and specific glucose values, or with limited attention given to overall glycemic control. This leaves a gap in expanding the scope of digital biomarkers for overall glycemic control in diabetes management. To address such a research gap, we propose GluMarker -- an end-to-end framework for modeling digital biomarkers using broader factors sources to predict glycemic control. Through the assessment and refinement of various machine learning baselines, GluMarker achieves state-of-the-art on Anderson's dataset in predicting next-day glycemic control. Moreover, our research identifies key digital biomarkers for the next day's glycemic control prediction. These identified biomarkers are instrumental in illuminating the daily factors that influence glycemic management, offering vital insights for diabetes care.

cs.AI↗

arXiv2024

Toward Short-Term Glucose Prediction Solely Based on CGM Time Series

The global diabetes epidemic highlights the importance of maintaining good glycemic control. Glucose prediction is a fundamental aspect of diabetes management, facilitating real-time decision-making. Recent research has introduced models focusing on long-term glucose trend prediction, which are unsuitable for real-time decision-making and result in delayed responses. Conversely, models designed to respond to immediate glucose level changes cannot analyze glucose variability comprehensively. Moreover, contemporary research generally integrates various physiological parameters (e.g. insulin doses, food intake, etc.), which inevitably raises data privacy concerns. To bridge such a research gap, we propose TimeGlu -- an end-to-end pipeline for short-term glucose prediction solely based on CGM time series data. We implement four baseline methods to conduct a comprehensive comparative analysis of the model's performance. Through extensive experiments on two contrasting datasets (CGM Glucose and Colas dataset), TimeGlu achieves state-of-the-art performance without the need for additional personal data from patients, providing effective guidance for real-world diabetic glucose management.

cs.AI↗

arXiv2024

CrossGP: Cross-Day Glucose Prediction Excluding Physiological Information

The increasing number of diabetic patients is a serious issue in society today, which has significant negative impacts on people's health and the country's financial expenditures. Because diabetes may develop into potential serious complications, early glucose prediction for diabetic patients is necessary for timely medical treatment. Existing glucose prediction methods typically utilize patients' private data (e.g. age, gender, ethnicity) and physiological parameters (e.g. blood pressure, heart rate) as reference features for glucose prediction, which inevitably leads to privacy protection concerns. Moreover, these models generally focus on either long-term (monthly-based) or short-term (minute-based) predictions. Long-term prediction methods are generally inaccurate because of the external uncertainties that can greatly affect the glucose values, while short-term ones fail to provide timely medical guidance. Based on the above issues, we propose CrossGP, a novel machine-learning framework for cross-day glucose prediction solely based on the patient's external activities without involving any physiological parameters. Meanwhile, we implement three baseline models for comparison. Extensive experiments on Anderson's dataset strongly demonstrate the superior performance of CrossGP and prove its potential for future real-life applications.

cs.AI↗

arXiv2024

Efflex: Efficient and Flexible Pipeline for Spatio-Temporal Trajectory Graph Modeling and Representation Learning

In the landscape of spatio-temporal data analytics, effective trajectory representation learning is paramount. To bridge the gap of learning accurate representations with efficient and flexible mechanisms, we introduce Efflex, a comprehensive pipeline for transformative graph modeling and representation learning of the large-volume spatio-temporal trajectories. Efflex pioneers the incorporation of a multi-scale k-nearest neighbors (KNN) algorithm with feature fusion for graph construction, marking a leap in dimensionality reduction techniques by preserving essential data features. Moreover, the groundbreaking graph construction mechanism and the high-performance lightweight GCN increase embedding extraction speed by up to 36 times faster. We further offer Efflex in two versions, Efflex-L for scenarios demanding high accuracy, and Efflex-B for environments requiring swift data processing. Comprehensive experimentation with the Porto and Geolife datasets validates our approach, positioning Efflex as the state-of-the-art in the domain. Such enhancements in speed and accuracy highlight the versatility of Efflex, underscoring its wide-ranging potential for deployment in time-sensitive and computationally constrained applications.

cs.LG↗