arXiv subjects

Yan Lu

Publications and source records attributed to Yan Lu.

At least 145 records · Page 8Linked to original sources

arXiv2025

StreamGS: Online Generalizable Gaussian Splatting Reconstruction for Unposed Image Streams

The advent of 3D Gaussian Splatting (3DGS) has advanced 3D scene reconstruction and novel view synthesis. With the growing interest of interactive applications that need immediate feedback, online 3DGS reconstruction in real-time is in high demand. However, none of existing methods yet meet the demand due to three main challenges: the absence of predetermined camera parameters, the need for generalizable 3DGS optimization, and the necessity of reducing redundancy. We propose StreamGS, an online generalizable 3DGS reconstruction method for unposed image streams, which progressively transform image streams to 3D Gaussian streams by predicting and aggregating per-frame Gaussians. Our method overcomes the limitation of the initial point reconstruction \cite{dust3r} in tackling out-of-domain (OOD) issues by introducing a content adaptive refinement. The refinement enhances cross-frame consistency by establishing reliable pixel correspondences between adjacent frames. Such correspondences further aid in merging redundant Gaussians through cross-frame feature aggregation. The density of Gaussians is thereby reduced, empowering online reconstruction by significantly lowering computational and memory costs. Extensive experiments on diverse datasets have demonstrated that StreamGS achieves quality on par with optimization-based approaches but does so 150 times faster, and exhibits superior generalizability in handling OOD scenes.

cs.CV↗

arXiv2025

Towards Practical Real-Time Neural Video Compression

We introduce a practical real-time neural video codec (NVC) designed to deliver high compression ratio, low latency and broad versatility. In practice, the coding speed of NVCs depends on 1) computational costs, and 2) non-computational operational costs, such as memory I/O and the number of function calls. While most efficient NVCs prioritize reducing computational cost, we identify operational cost as the primary bottleneck to achieving higher coding speed. Leveraging this insight, we introduce a set of efficiency-driven design improvements focused on minimizing operational costs. Specifically, we employ implicit temporal modeling to eliminate complex explicit motion modules, and use single low-resolution latent representations rather than progressive downsampling. These innovations significantly accelerate NVC without sacrificing compression quality. Additionally, we implement model integerization for consistent cross-device coding and a module-bank-based rate control scheme to improve practical adaptability. Experiments show our proposed DCVC-RT achieves an impressive average encoding/decoding speed at 125.2/112.8 fps (frames per second) for 1080p video, while saving an average of 21% in bitrate compared to H.266/VTM. The code is available at https://github.com/microsoft/DCVC.

eess.IV↗

arXiv2025

UVRM: A Scalable 3D Reconstruction Model from Unposed Videos

Large Reconstruction Models (LRMs) have recently become a popular method for creating 3D foundational models. Training 3D reconstruction models with 2D visual data traditionally requires prior knowledge of camera poses for the training samples, a process that is both time-consuming and prone to errors. Consequently, 3D reconstruction training has been confined to either synthetic 3D datasets or small-scale datasets with annotated poses. In this study, we investigate the feasibility of 3D reconstruction using unposed video data of various objects. We introduce UVRM, a novel 3D reconstruction model capable of being trained and evaluated on monocular videos without requiring any information about the pose. UVRM uses a transformer network to implicitly aggregate video frames into a pose-invariant latent feature space, which is then decoded into a tri-plane 3D representation. To obviate the need for ground-truth pose annotations during training, UVRM employs a combination of the score distillation sampling (SDS) method and an analysis-by-synthesis approach, progressively synthesizing pseudo novel-views using a pre-trained diffusion model. We qualitatively and quantitatively evaluate UVRM's performance on the G-Objaverse and CO3D datasets without relying on pose information. Extensive experiments show that UVRM is capable of effectively and efficiently reconstructing a wide range of 3D objects from unposed videos.

cs.CV↗

arXiv2025

UniGraspTransformer: Simplified Policy Distillation for Scalable Dexterous Robotic Grasping

We introduce UniGraspTransformer, a universal Transformer-based network for dexterous robotic grasping that simplifies training while enhancing scalability and performance. Unlike prior methods such as UniDexGrasp++, which require complex, multi-step training pipelines, UniGraspTransformer follows a streamlined process: first, dedicated policy networks are trained for individual objects using reinforcement learning to generate successful grasp trajectories; then, these trajectories are distilled into a single, universal network. Our approach enables UniGraspTransformer to scale effectively, incorporating up to 12 self-attention blocks for handling thousands of objects with diverse poses. Additionally, it generalizes well to both idealized and real-world inputs, evaluated in state-based and vision-based settings. Notably, UniGraspTransformer generates a broader range of grasping poses for objects in various shapes and orientations, resulting in more diverse grasp strategies. Experimental results demonstrate significant improvements over state-of-the-art, UniDexGrasp++, across various object categories, achieving success rate gains of 3.5%, 7.7%, and 10.1% on seen objects, unseen objects within seen categories, and completely unseen objects, respectively, in the vision-based setting. Project page: https://dexhand.github.io/UniGraspTransformer.

cs.RO↗

arXiv2025

GistVis: Automatic Generation of Word-scale Visualizations from Data-rich Documents

Data-rich documents are ubiquitous in various applications, yet they often rely solely on textual descriptions to convey data insights. Prior research primarily focused on providing visualization-centric augmentation to data-rich documents. However, few have explored using automatically generated word-scale visualizations to enhance the document-centric reading process. As an exploratory step, we propose GistVis, an automatic pipeline that extracts and visualizes data insight from text descriptions. GistVis decomposes the generation process into four modules: Discoverer, Annotator, Extractor, and Visualizer, with the first three modules utilizing the capabilities of large language models and the fourth using visualization design knowledge. Technical evaluation including a comparative study on Discoverer and an ablation study on Annotator reveals decent performance of GistVis. Meanwhile, the user study (N=12) showed that GistVis could generate satisfactory word-scale visualizations, indicating its effectiveness in facilitating users' understanding of data-rich documents (+5.6% accuracy) while significantly reducing their mental demand (p=0.016) and perceived effort (p=0.033).

cs.HC↗

arXiv2025

GUPNet++: Geometry Uncertainty Propagation Network for Monocular 3D Object Detection

Geometry plays a significant role in monocular 3D object detection. It can be used to estimate object depth by using the perspective projection between object's physical size and 2D projection in the image plane, which can introduce mathematical priors into deep models. However, this projection process also introduces error amplification, where the error of the estimated height is amplified and reflected into the projected depth. It leads to unreliable depth inferences and also impairs training stability. To tackle this problem, we propose a novel Geometry Uncertainty Propagation Network (GUPNet++) by modeling geometry projection in a probabilistic manner. This ensures depth predictions are well-bounded and associated with a reasonable uncertainty. The significance of introducing such geometric uncertainty is two-fold: (1). It models the uncertainty propagation relationship of the geometry projection during training, improving the stability and efficiency of the end-to-end model learning. (2). It can be derived to a highly reliable confidence to indicate the quality of the 3D detection result, enabling more reliable detection inference. Experiments show that the proposed approach not only obtains (state-of-the-art) SOTA performance in image-based monocular 3D detection but also demonstrates superiority in efficacy with a simplified framework.

cs.CV↗

arXiv2024

GSemSplat: Generalizable Semantic 3D Gaussian Splatting from Uncalibrated Image Pairs

Modeling and understanding the 3D world is crucial for various applications, from augmented reality to robotic navigation. Recent advancements based on 3D Gaussian Splatting have integrated semantic information from multi-view images into Gaussian primitives. However, these methods typically require costly per-scene optimization from dense calibrated images, limiting their practicality. In this paper, we consider the new task of generalizable 3D semantic field modeling from sparse, uncalibrated image pairs. Building upon the Splatt3R architecture, we introduce GSemSplat, a framework that learns open-vocabulary semantic representations linked to 3D Gaussians without the need for per-scene optimization, dense image collections or calibration. To ensure effective and reliable learning of semantic features in 3D space, we employ a dual-feature approach that leverages both region-specific and context-aware semantic features as supervision in the 2D space. This allows us to capitalize on their complementary strengths. Experimental results on the ScanNet++ dataset demonstrate the effectiveness and superiority of our approach compared to the traditional scene-specific method. We hope our work will inspire more research into generalizable 3D understanding.

cs.CV↗

arXiv2024

SLAM-Omni: Timbre-Controllable Voice Interaction System with Single-Stage Training

Recent advancements highlight the potential of end-to-end real-time spoken dialogue systems, showcasing their low latency and high quality. In this paper, we introduce SLAM-Omni, a timbre-controllable, end-to-end voice interaction system with single-stage training. SLAM-Omni achieves zero-shot timbre control by modeling spoken language with semantic tokens and decoupling speaker information to a vocoder. By predicting grouped speech semantic tokens at each step, our method significantly reduces the sequence length of audio tokens, accelerating both training and inference. Additionally, we propose historical text prompting to compress dialogue history, facilitating efficient multi-round interactions. Comprehensive evaluations reveal that SLAM-Omni outperforms prior models of similar scale, requiring only 15 hours of training on 4 GPUs with limited data. Notably, it is the first spoken dialogue system to achieve competitive performance with a single-stage training approach, eliminating the need for pre-training on TTS or ASR tasks. Further experiments validate its multilingual and multi-turn dialogue capabilities on larger datasets.

eess.AS↗

arXiv2024

MEATRD: Multimodal Anomalous Tissue Region Detection Enhanced with Spatial Transcriptomics

The detection of anomalous tissue regions (ATRs) within affected tissues is crucial in clinical diagnosis and pathological studies. Conventional automated ATR detection methods, primarily based on histology images alone, falter in cases where ATRs and normal tissues have subtle visual differences. The recent spatial transcriptomics (ST) technology profiles gene expressions across tissue regions, offering a molecular perspective for detecting ATRs. However, there is a dearth of ATR detection methods that effectively harness complementary information from both histology images and ST. To address this gap, we propose MEATRD, a novel ATR detection method that integrates histology image and ST data. MEATRD is trained to reconstruct image patches and gene expression profiles of normal tissue spots (inliers) from their multimodal embeddings, followed by learning a one-class classification AD model based on latent multimodal reconstruction errors. This strategy harmonizes the strengths of reconstruction-based and one-class classification approaches. At the heart of MEATRD is an innovative masked graph dual-attention transformer (MGDAT) network, which not only facilitates cross-modality and cross-node information sharing but also addresses the model over-generalization issue commonly seen in reconstruction-based AD methods. Additionally, we demonstrate that modality-specific, task-relevant information is collated and condensed in multimodal bottleneck encoding generated in MGDAT, marking the first theoretical analysis of the informational properties of multimodal bottleneck encoding. Extensive evaluations across eight real ST datasets reveal MEATRD's superior performance in ATR detection, surpassing various state-of-the-art AD methods. Remarkably, MEATRD also proves adept at discerning ATRs that only show slight visual deviations from normal tissues.

cs.CV↗

arXiv2024

The classification of real and bogus transients using active learning and semi-supervised learning

Deep-learning-based methods have been favored in astrophysics owing to their adaptability and remarkable performance and have been applied to the task of the classification of real and bogus transients. Different from most existing approaches which necessitate massive yet expensive annotated data, We aim to leverage training samples with only 1000 labels available to discover real sources that vary in brightness over time in the early stage of the WFST 6-year survey. Methods. We present a novel deep-learning method that combines active learning and semi-supervised learning to construct a competitive real/bogus classifier. Our method incorporates an active learning stage, where we actively select the most informative or uncertain samples for annotation. This stage aims to achieve higher model performance by leveraging fewer labeled samples, thus reducing annotation costs and improving the overall learning process efficiency. Furthermore, our approach involves a semi-supervised learning stage that exploits the unlabeled data to enhance the model's performance and achieve superior results compared to using only the limited labeled data.

astro-ph.IM↗

arXiv2024

Real-time surface plasmon polariton propagation in silver nanowires

Electron microscopy techniques such as electron energy-loss spectroscopy (EELS) facilitate the spatio-spectral characterization of plasmonic nanostructures. In this work, a time-dependent perspective is presented, which significantly enhances the utility of EELS. Specifically, silver nanowires offer the material and geometric features for various high-quality plasmonic excitations. This provides an ideal illustrative system for combined experimental-theoretical analyses of the different plasmonic excitations and their real-time dynamics. It is demonstrated how the plasmonic excitations propagating inside the wire repeatedly interact with the swift electrons in an EELS configuration. In addition, the role of azimuthal modes, often overlooked for very thin wires, is observed and analyzed in both the energy-loss spectrum and the dynamical perspective. Such a complete understanding of the interaction of electrons and plasmonic excitation is key for the design of efficient plasmonic sensors, the study of hot electron dynamics in metals, and applications in the context of electron quantum optics, where full control of the spatial and temporal characteristics of the fields at the nanometer and femtosecond scales is highly desirable.

physics.optics↗

arXiv2024

Towards reproducible machine learning-based process monitoring and quality prediction research for additive manufacturing

Machine learning (ML)-based cyber-physical systems (CPSs) have been extensively developed to improve the print quality of additive manufacturing (AM). However, the reproducibility of these systems, as presented in published research, has not been thoroughly investigated due to a lack of formal evaluation methods. Reproducibility, a critical component of trustworthy artificial intelligence, is achieved when an independent team can replicate the findings or artifacts of a study using a different experimental setup and achieve comparable performance. In many publications, critical information necessary for reproduction is often missing, resulting in systems that fail to replicate the reported performance. This paper proposes a reproducibility investigation pipeline and a reproducibility checklist for ML-based process monitoring and quality prediction systems for AM. The pipeline guides researchers through the key steps required to reproduce a study, while the checklist systematically extracts reproducibility-relevant information from the publication. We validated the proposed approach through two case studies: reproducing a fused filament fabrication warping detection system and a laser powder bed fusion melt pool area prediction model. Both case studies confirmed that the pipeline and checklist successfully identified missing information, improved reproducibility, and enhanced the performance of reproduced systems. Based on the proposed checklist, a reproducibility survey was conducted to assess the current reproducibility status within this research domain. By addressing this research gap, the proposed methods aim to enhance trustworthiness and rigor in ML-based AM research, with potential applicability to other ML-based CPSs.

cs.CE↗

arXiv2024

Adaptive high-precision sound source localization at low frequencies based on convolutional neural network

Sound source localization (SSL) technology plays a crucial role in various application areas such as fault diagnosis, speech separation, and vibration noise reduction. Although beamforming algorithms are widely used in SSL, their resolution at low frequencies is limited. In recent years, deep learning-based SSL methods have significantly improved their accuracy by employing large microphone arrays and training case specific neural networks, however, this could lead to narrow applicability. To address these issues, this paper proposes a convolutional neural network-based method for high-precision SSL, which is adaptive in the lower frequency range under 1kHz with varying numbers of sound sources and microphone array-to-scanning grid distances. It takes the pressure distribution on a relatively small microphone array as input to the neural network, and employs customized training labels and loss function to train the model. Prediction accuracy, adaptability and robustness of the trained model under certain signal-to-noise ratio (SNR) are evaluated using randomly generated test datasets, and compared with classical beamforming algorithms, CLEAN-SC and DAMAS. Results of both planar and spatial sound source distributions show that the proposed neural network model significantly improves low-frequency localization accuracy, demonstrating its effectiveness and potential in SSL.

cs.SD↗

arXiv2024

Investigation on domain adaptation of additive manufacturing monitoring systems to enhance digital twin reusability

Powder bed fusion (PBF) is an emerging metal additive manufacturing (AM) technology that enables rapid fabrication of complex geometries. However, defects such as pores and balling may occur and lead to structural unconformities, thus compromising the mechanical performance of the part. This has become a critical challenge for quality assurance as the nature of some defects is stochastic during the process and invisible from the exterior. To address this issue, digital twin (DT) using machine learning (ML)-based modeling can be deployed for AM process monitoring and control. Melt pool is one of the most commonly observed physical phenomena for process monitoring, usually by high-speed cameras. Once labeled and preprocessed, the melt pool images are used to train ML-based models for DT applications such as process anomaly detection and print quality evaluation. Nonetheless, the reusability of DTs is restricted due to the wide variability of AM settings, including AM machines and monitoring instruments. The performance of the ML models trained using the dataset collected from one setting is usually compromised when applied to other settings. This paper proposes a knowledge transfer pipeline between different AM settings to enhance the reusability of AM DTs. The source and target datasets are collected from the National Institute of Standards and Technology and National Cheng Kung University with different cameras, materials, AM machines, and process parameters. The proposed pipeline consists of four steps: data preprocessing, data augmentation, domain alignment, and decision alignment. Compared with the model trained only using the source dataset, this pipeline increased the melt pool anomaly detection accuracy by 31% without any labeled training data from the target dataset.

cs.CE↗

arXiv2024

UWF-RI2FA: Generating Multi-frame Ultrawide-field Fluorescein Angiography from Ultrawide-field Retinal Imaging Improves Diabetic Retinopathy Stratification

Ultrawide-field fluorescein angiography (UWF-FA) facilitates diabetic retinopathy (DR) detection by providing a clear visualization of peripheral retinal lesions. However, the intravenous dye injection with potential risks hamper its application. We aim to acquire dye-free UWF-FA images from noninvasive UWF retinal imaging (UWF-RI) using generative artificial intelligence (GenAI) and evaluate its effectiveness in DR screening. A total of 18,321 UWF-FA images of different phases were registered with corresponding UWF-RI images and fed into a generative adversarial networks (GAN)-based model for training. The quality of generated UWF-FA images was evaluated through quantitative metrics and human evaluation. The DeepDRiD dataset was used to externally assess the contribution of generated UWF-FA images to DR classification, using area under the receiver operating characteristic curve (AUROC) as outcome metrics. The generated early, mid, and late phase UWF-FA images achieved high authenticity, with multi-scale similarity scores ranging from 0.70 to 0.91 and qualitative visual scores ranging from 1.64 to 1.98 (1=real UWF-FA quality). In fifty randomly selected images, 56% to 76% of the generated images were difficult to distinguish from real images in the Turing test. Moreover, adding these generated UWF-FA images for DR classification significantly increased the AUROC from 0.869 to 0.904 compared to the baseline model using UWF-RI images (P < .001). The model successfully generates realistic multi-frame UWF-FA images for enhancing DR stratification without intravenous dye injection.

eess.IV↗

arXiv2024

Convert and Speak: Zero-shot Accent Conversion with Minimum Supervision

Low resource of parallel data is the key challenge of accent conversion(AC) problem in which both the pronunciation units and prosody pattern need to be converted. We propose a two-stage generative framework "convert-and-speak" in which the conversion is only operated on the semantic token level and the speech is synthesized conditioned on the converted semantic token with a speech generative model in target accent domain. The decoupling design enables the "speaking" module to use massive amount of target accent speech and relieves the parallel data required for the "conversion" module. Conversion with the bridge of semantic token also relieves the requirement for the data with text transcriptions and unlocks the usage of language pre-training technology to further efficiently reduce the need of parallel accent speech data. To reduce the complexity and latency of "speaking", a single-stage AR generative model is designed to achieve good quality as well as lower computation cost. Experiments on Indian-English to general American-English conversion show that the proposed framework achieves state-of-the-art performance in accent similarity, speech quality, and speaker maintenance with only 15 minutes of weakly parallel data which is not constrained to the same speaker. Extensive experimentation with diverse accent types suggests that this framework possesses a high degree of adaptability, making it readily scalable to accommodate other accents with low-resource data. Audio samples are available at https://www.microsoft.com/en-us/research/project/convert-and-speak-zero-shot-accent-conversion-with-minimumsupervision/.

cs.SD↗

arXiv2024

Diffusion Model with Cross Attention as an Inductive Bias for Disentanglement

Disentangled representation learning strives to extract the intrinsic factors within observed data. Factorizing these representations in an unsupervised manner is notably challenging and usually requires tailored loss functions or specific structural designs. In this paper, we introduce a new perspective and framework, demonstrating that diffusion models with cross-attention can serve as a powerful inductive bias to facilitate the learning of disentangled representations. We propose to encode an image to a set of concept tokens and treat them as the condition of the latent diffusion for image reconstruction, where cross-attention over the concept tokens is used to bridge the interaction between the encoder and diffusion. Without any additional regularization, this framework achieves superior disentanglement performance on the benchmark datasets, surpassing all previous methods with intricate designs. We have conducted comprehensive ablation studies and visualization analysis, shedding light on the functioning of this model. This is the first work to reveal the potent disentanglement capability of diffusion models with cross-attention, requiring no complex designs. We anticipate that our findings will inspire more investigation on exploring diffusion for disentangled representation learning towards more sophisticated data analysis and understanding.

cs.CV↗

arXiv2024

Breaking through the learning plateaus of in-context learning in Transformer

In-context learning, i.e., learning from context examples, is an impressive ability of Transformer. Training Transformers to possess this in-context learning skill is computationally intensive due to the occurrence of learning plateaus, which are periods within the training process where there is minimal or no enhancement in the model's in-context learning capability. To study the mechanism behind the learning plateaus, we conceptually seperate a component within the model's internal representation that is exclusively affected by the model's weights. We call this the "weights component", and the remainder is identified as the "context component". By conducting meticulous and controlled experiments on synthetic tasks, we note that the persistence of learning plateaus correlates with compromised functionality of the weights component. Recognizing the impaired performance of the weights component as a fundamental behavior drives learning plateaus, we have developed three strategies to expedite the learning of Transformers. The effectiveness of these strategies is further confirmed in natural language processing tasks. In conclusion, our research demonstrates the feasibility of cultivating a powerful in-context learning ability within AI systems in an eco-friendly manner.

cs.LG↗