arXiv subjects

Yu Liu

Publications and source records attributed to Yu Liu.

At least 577 records · Page 32Linked to original sources

arXiv2024

Render-and-Compare: Cross-View 6 DoF Localization from Noisy Prior

Despite the significant progress in 6-DoF visual localization, researchers are mostly driven by ground-level benchmarks. Compared with aerial oblique photography, ground-level map collection lacks scalability and complete coverage. In this work, we propose to go beyond the traditional ground-level setting and exploit the cross-view localization from aerial to ground. We solve this problem by formulating camera pose estimation as an iterative render-and-compare pipeline and enhancing the robustness through augmenting seeds from noisy initial priors. As no public dataset exists for the studied problem, we collect a new dataset that provides a variety of cross-view images from smartphones and drones and develop a semi-automatic system to acquire ground-truth poses for query images. We benchmark our method as well as several state-of-the-art baselines and demonstrate that our method outperforms other approaches by a large margin.

cs.CV↗

arXiv2024

Incremental Multiview Point Cloud Registration

In this paper, we present a novel approach for multiview point cloud registration. Different from previous researches that typically employ a global scheme for multiview registration, we propose to adopt an incremental pipeline to progressively align scans into a canonical coordinate system. Specifically, drawing inspiration from image-based 3D reconstruction, our approach first builds a sparse scan graph with scan retrieval and geometric verification. Then, we perform incremental registration via initialization, next scan selection and registration, Track create and continue, and Bundle Adjustment. Additionally, for detector-free matchers, we incorporate a Track refinement process. This process primarily constructs a coarse multiview registration and refines the model by adjusting the positions of the keypoints on the Track. Experiments demonstrate that the proposed framework outperforms existing multiview registration methods on three benchmark datasets. The code is available at https://github.com/Choyaa/IncreMVR.

cs.CV↗

arXiv2024

A Study on Activity Visualization for Smart Watches

This paper investigates the use of visualization to display activity data on smartwatches by surveying the data visual presentations proposed by 80 smartwatch models currently available on the Chinese e-commerce platform JD.com and, later, surveying the preferences of 41 users concerning these visualizations. The results show that despite radial bar charts are the most popular visualization for activity data on smartwatches, the users' preferences might be influenced by their familiarity with these charts. These findings from this survey will be valuable for designers, developers, and researchers who are interested in creating innovative and effective solutions for activity visualization on smartwatches.

cs.HC↗

arXiv2024

MM-Instruct: Generated Visual Instructions for Large Multimodal Model Alignment

This paper introduces MM-Instruct, a large-scale dataset of diverse and high-quality visual instruction data designed to enhance the instruction-following capabilities of large multimodal models (LMMs). While existing visual instruction datasets often focus on question-answering, they struggle to generalize to broader application scenarios such as creative writing, summarization, or image analysis. To address these limitations, we propose a novel approach to constructing MM-Instruct that leverages the strong instruction-following capabilities of existing LLMs to generate novel visual instruction data from large-scale but conventional image captioning datasets. MM-Instruct first leverages ChatGPT to automatically generate diverse instructions from a small set of seed instructions through augmenting and summarization. It then matches these instructions with images and uses an open-sourced large language model (LLM) to generate coherent answers to the instruction-image pairs. The LLM is grounded by the detailed text descriptions of images in the whole answer generation process to guarantee the alignment of the instruction data. Moreover, we introduce a benchmark based on the generated instruction data to evaluate the instruction-following capabilities of existing LMMs. We demonstrate the effectiveness of MM-Instruct by training a LLaVA-1.5 model on the generated data, denoted as LLaVA-Instruct, which exhibits significant improvements in instruction-following capabilities compared to LLaVA-1.5 models. The MM-Instruct dataset, benchmark, and pre-trained models are available at https://github.com/jihaonew/MM-Instruct.

cs.CV↗

arXiv2024

Continuous 3D Myocardial Motion Tracking via Echocardiography

Myocardial motion tracking stands as an essential clinical tool in the prevention and detection of cardiovascular diseases (CVDs), the foremost cause of death globally. However, current techniques suffer from incomplete and inaccurate motion estimation of the myocardium in both spatial and temporal dimensions, hindering the early identification of myocardial dysfunction. To address these challenges, this paper introduces the Neural Cardiac Motion Field (NeuralCMF). NeuralCMF leverages implicit neural representation (INR) to model the 3D structure and the comprehensive 6D forward/backward motion of the heart. This method surpasses pixel-wise limitations by offering the capability to continuously query the precise shape and motion of the myocardium at any specific point throughout the cardiac cycle, enhancing the detailed analysis of cardiac dynamics beyond traditional speckle tracking. Notably, NeuralCMF operates without the need for paired datasets, and its optimization is self-supervised through the physics knowledge priors in both space and time dimensions, ensuring compatibility with both 2D and 3D echocardiogram video inputs. Experimental validations across three representative datasets support the robustness and innovative nature of the NeuralCMF, marking significant advantages over existing state-of-the-art methods in cardiac imaging and motion tracking.

eess.IV↗

arXiv2024

GMP-AR: Granularity Message Passing and Adaptive Reconciliation for Temporal Hierarchy Forecasting

Time series forecasts of different temporal granularity are widely used in real-world applications, e.g., sales prediction in days and weeks for making different inventory plans. However, these tasks are usually solved separately without ensuring coherence, which is crucial for aligning downstream decisions. Previous works mainly focus on ensuring coherence with some straightforward methods, e.g., aggregation from the forecasts of fine granularity to the coarse ones, and allocation from the coarse granularity to the fine ones. These methods merely take the temporal hierarchical structure to maintain coherence without improving the forecasting accuracy. In this paper, we propose a novel granularity message-passing mechanism (GMP) that leverages temporal hierarchy information to improve forecasting performance and also utilizes an adaptive reconciliation (AR) strategy to maintain coherence without performance loss. Furthermore, we introduce an optimization module to achieve task-based targets while adhering to more real-world constraints. Experiments on real-world datasets demonstrate that our framework (GMP-AR) achieves superior performances on temporal hierarchical forecasting tasks compared to state-of-the-art methods. In addition, our framework has been successfully applied to a real-world task of payment traffic management in Alipay by integrating with the task-based optimization module.

cs.LG↗

arXiv2024

From Pixels to Progress: Generating Road Network from Satellite Imagery for Socioeconomic Insights in Impoverished Areas

The Sustainable Development Goals (SDGs) aim to resolve societal challenges, such as eradicating poverty and improving the lives of vulnerable populations in impoverished areas. Those areas rely on road infrastructure construction to promote accessibility and economic development. Although publicly available data like OpenStreetMap is available to monitor road status, data completeness in impoverished areas is limited. Meanwhile, the development of deep learning techniques and satellite imagery shows excellent potential for earth monitoring. To tackle the challenge of road network assessment in impoverished areas, we develop a systematic road extraction framework combining an encoder-decoder architecture and morphological operations on satellite imagery, offering an integrated workflow for interdisciplinary researchers. Extensive experiments of road network extraction on real-world data in impoverished regions achieve a 42.7% enhancement in the F1-score over the baseline methods and reconstruct about 80% of the actual roads. We also propose a comprehensive road network dataset covering approximately 794,178 km2 area and 17.048 million people in 382 impoverished counties in China. The generated dataset is further utilized to conduct socioeconomic analysis in impoverished counties, showing that road network construction positively impacts regional economic development. The technical appendix, code, and generated dataset can be found at https://github.com/tsinghua-fib-lab/Road_network_extraction_impoverished_counties.

cs.CV↗

arXiv2024

Enhancing Vision-Language Model with Unmasked Token Alignment

Contrastive pre-training on image-text pairs, exemplified by CLIP, becomes a standard technique for learning multi-modal visual-language representations. Although CLIP has demonstrated remarkable performance, training it from scratch on noisy web-scale datasets is computationally demanding. On the other hand, mask-then-predict pre-training approaches, like Masked Image Modeling (MIM), offer efficient self-supervised learning for single-modal representations. This paper introduces Unmasked Token Alignment (UTA), a method that leverages existing CLIP models to further enhance its vision-language representations. UTA trains a Vision Transformer (ViT) by aligning unmasked visual tokens to the corresponding image tokens from a frozen CLIP vision encoder, which automatically aligns the ViT model with the CLIP text encoder. The pre-trained ViT can be directly applied for zero-shot evaluation even without training on image-text pairs. Compared to MIM approaches, UTA does not suffer from training-finetuning inconsistency and is much more training-efficient by avoiding using the extra [MASK] tokens. Extensive experimental results demonstrate that UTA can enhance CLIP models and outperform existing MIM methods on various uni- and multi-modal benchmarks. Code and models are available at https://github.com/jihaonew/UTA.

cs.CV↗

arXiv2024

Compressed Video Quality Enhancement with Temporal Group Alignment and Fusion

In this paper, we propose a temporal group alignment and fusion network to enhance the quality of compressed videos by using the long-short term correlations between frames. The proposed model consists of the intra-group feature alignment (IntraGFA) module, the inter-group feature fusion (InterGFF) module, and the feature enhancement (FE) module. We form the group of pictures (GoP) by selecting frames from the video according to their temporal distances to the target enhanced frame. With this grouping, the composed GoP can contain either long- or short-term correlated information of neighboring frames. We design the IntraGFA module to align the features of frames of each GoP to eliminate the motion existing between frames. We construct the InterGFF module to fuse features belonging to different GoPs and finally enhance the fused features with the FE module to generate high-quality video frames. The experimental results show that our proposed method achieves up to 0.05dB gain and lower complexity compared to the state-of-the-art method.

cs.CV↗

arXiv2024

The Devil is in the Neurons: Interpreting and Mitigating Social Biases in Pre-trained Language Models

Pre-trained Language models (PLMs) have been acknowledged to contain harmful information, such as social biases, which may cause negative social impacts or even bring catastrophic results in application. Previous works on this problem mainly focused on using black-box methods such as probing to detect and quantify social biases in PLMs by observing model outputs. As a result, previous debiasing methods mainly finetune or even pre-train language models on newly constructed anti-stereotypical datasets, which are high-cost. In this work, we try to unveil the mystery of social bias inside language models by introducing the concept of {\sc Social Bias Neurons}. Specifically, we propose {\sc Integrated Gap Gradients (IG$^2$)} to accurately pinpoint units (i.e., neurons) in a language model that can be attributed to undesirable behavior, such as social bias. By formalizing undesirable behavior as a distributional property of language, we employ sentiment-bearing prompts to elicit classes of sensitive words (demographics) correlated with such sentiments. Our IG$^2$ thus attributes the uneven distribution for different demographics to specific Social Bias Neurons, which track the trail of unwanted behavior inside PLM units to achieve interoperability. Moreover, derived from our interpretable technique, {\sc Bias Neuron Suppression (BNS)} is further proposed to mitigate social biases. By studying BERT, RoBERTa, and their attributable differences from debiased FairBERTa, IG$^2$ allows us to locate and suppress identified neurons, and further mitigate undesired behaviors. As measured by prior metrics from StereoSet, our model achieves a higher degree of fairness while maintaining language modeling ability with low cost.

cs.CL↗

arXiv2024

Zero-shot Image Editing with Reference Imitation

Image editing serves as a practical yet challenging task considering the diverse demands from users, where one of the hardest parts is to precisely describe how the edited image should look like. In this work, we present a new form of editing, termed imitative editing, to help users exercise their creativity more conveniently. Concretely, to edit an image region of interest, users are free to directly draw inspiration from some in-the-wild references (e.g., some relative pictures come across online), without having to cope with the fit between the reference and the source. Such a design requires the system to automatically figure out what to expect from the reference to perform the editing. For this purpose, we propose a generative training framework, dubbed MimicBrush, which randomly selects two frames from a video clip, masks some regions of one frame, and learns to recover the masked regions using the information from the other frame. That way, our model, developed from a diffusion prior, is able to capture the semantic correspondence between separate images in a self-supervised manner. We experimentally show the effectiveness of our method under various test cases as well as its superiority over existing alternatives. We also construct a benchmark to facilitate further research.

cs.CV↗

arXiv2024

Self-supervised Representations and Node Embedding Graph Neural Networks for Accurate and Multi-scale Analysis of Materials

Supervised machine learning algorithms, such as graph neural networks (GNN), have successfully predicted material properties. However, the superior performance of GNN usually relies on end-to-end learning on large material datasets, which may lose the physical insight of multi-scale information about materials. And the process of labeling data consumes many resources and inevitably introduces errors, which constrains the accuracy of prediction. We propose to train the GNN model by self-supervised learning on the node and edge information of the crystal graph. Compared with the popular manually constructed material descriptors, the self-supervised atomic representation can reach better prediction performance on material properties. Furthermore, it may provide physical insights by tuning the range information. Applying the self-supervised atomic representation on the magnetic moment datasets, we show how they can extract rules and information from the magnetic materials. To incorporate rich physical information into the GNN model, we develop the node embedding graph neural networks (NEGNN) framework and show significant improvements in the prediction performance. The self-supervised material representation and the NEGNN framework may investigate in-depth information from materials and can be applied to small datasets with increased prediction accuracy.

cond-mat.mtrl-sci↗

arXiv2024

Anisotropic hybridization in CeRhSn

The optical conductivity $σ(ω,T)$ of CeRhSn was studied by broadband infrared spectroscopy. Temperature-dependent spectral weight transfer occurs over high energy ($0.8\,$eV) and temperature (${\sim}500\,$K) scales, classifying CeRhSn as a mixed valent compound. The optical conductivity reveals a substantial anisotropy in the electronic structure. Renormalization of $σ(ω,T)$ occurs as a function of temperature to a coherent Kondo state with concomitant effective mass generation. Associated spectroscopic signatures were reproduced remarkably well by the combination of density functional theory and dynamical mean field theory using a momentum-independent self energy. The theory shows that the anisotropy for energies $>10\,$meV is mainly driven by the bare three-dimensional electronic structure that is renormalized by local electronic correlations. The possible influence of magnetic frustration and quantum criticality is restricted to lower energies.

cond-mat.str-el↗

arXiv2024

GAN Inversion for Image Editing via Unsupervised Domain Adaptation

Existing GAN inversion methods work brilliantly in reconstructing high-quality (HQ) images while struggling with more common low-quality (LQ) inputs in practical application. To address this issue, we propose Unsupervised Domain Adaptation (UDA) in the inversion process, namely UDA-inversion, for effective inversion and editing of both HQ and LQ images. Regarding unpaired HQ images as the source domain and LQ images as the unlabeled target domain, we introduce a theoretical guarantee: loss value in the target domain is upper-bounded by loss in the source domain and a novel discrepancy function measuring the difference between two domains. Following that, we can only minimize this upper bound to obtain accurate latent codes for HQ and LQ images. Thus, constructive representations of HQ images can be spontaneously learned and transformed into LQ images without supervision. UDA-Inversion achieves a better PSNR of 22.14 on FFHQ dataset and performs comparably to supervised methods.

cs.CV↗

arXiv2024

Capturing dynamics and thermodynamics of a three-level quantum heat engine via programmable quantum circuits

This research employs the Kraus representation and Sz.-Nagy dilation theorem to model a three-level quantum heat on quantum circuits, investigating its dynamic evolution and thermodynamic performance. The feasibility of the dynamic model is validated by tracking the changes of population. On the basis of reinforcement learning algorithm, the optimal cycle of the quantum heat engine for maximal average power is proposed and verified by the thermodynamic model. The stability of quantum circuit simulations is scrutinized through a comparative analysis of theoretical and simulated results, predicated on an orthogonal test. These results affirm the practicality of simulating quantum heat engines on quantum circuits, offering potential for substantially curtailing the experimental expenses associated with the construction of such engines.

quant-ph↗

arXiv2024

DecisionNCE: Embodied Multimodal Representations via Implicit Preference Learning

Multimodal pretraining is an effective strategy for the trinity of goals of representation learning in autonomous robots: 1) extracting both local and global task progressions; 2) enforcing temporal consistency of visual representation; 3) capturing trajectory-level language grounding. Most existing methods approach these via separate objectives, which often reach sub-optimal solutions. In this paper, we propose a universal unified objective that can simultaneously extract meaningful task progression information from image sequences and seamlessly align them with language instructions. We discover that via implicit preferences, where a visual trajectory inherently aligns better with its corresponding language instruction than mismatched pairs, the popular Bradley-Terry model can transform into representation learning through proper reward reparameterizations. The resulted framework, DecisionNCE, mirrors an InfoNCE-style objective but is distinctively tailored for decision-making tasks, providing an embodied representation learning framework that elegantly extracts both local and global task progression features, with temporal consistency enforced through implicit time contrastive learning, while ensuring trajectory-level instruction grounding via multimodal joint encoding. Evaluation on both simulated and real robots demonstrates that DecisionNCE effectively facilitates diverse downstream policy learning tasks, offering a versatile solution for unified representation and reward learning. Project Page: https://2toinf.github.io/DecisionNCE/

cs.RO↗

arXiv2024

Integration of Scanning Probe Microscope with High-Performance Computing: fixed-policy and reward-driven workflows implementation

The rapid development of computation power and machine learning algorithms has paved the way for automating scientific discovery with a scanning probe microscope (SPM). The key elements towards operationalization of automated SPM are the interface to enable SPM control from Python codes, availability of high computing power, and development of workflows for scientific discovery. Here we build a Python interface library that enables controlling an SPM from either a local computer or a remote high-performance computer (HPC), which satisfies the high computation power need of machine learning algorithms in autonomous workflows. We further introduce a general platform to abstract the operations of SPM in scientific discovery into fixed-policy or reward-driven workflows. Our work provides a full infrastructure to build automated SPM workflows for both routine operations and autonomous scientific discovery with machine learning.

cond-mat.mtrl-sci↗

arXiv2024

Experimental demonstration of steady-state dynamics of three-level quantum heat engine using superconducting quantum circuits

The three-level system represents the smallest quantum system capable of autonomous cycling in quantum heat engines. This study proposes a method to simulate the steady-state dynamics of a three-level quantum heat engine by designing and implementing superconducting quantum circuits. Following error mitigation, the outcomes from the quantum circuit model designed in this study, when executed on a real quantum device, closely align with theoretical predictions, thereby validating the effectiveness of the circuit model. This study offers a novel approach for investigating three-level quantum heat engines, enabling the verification of theoretical research findings while also reducing the complexity and cost of experimental procedures.

quant-ph↗