arXiv subjects

Peng Lu

Publications and source records attributed to Peng Lu.

At least 91 records · Page 5Linked to original sources

arXiv2024

Draft on the Fly: Adaptive Self-Speculative Decoding using Cosine Similarity

We present a simple on the fly method for faster inference of large language models. Unlike other (self-)speculative decoding techniques, our method does not require fine-tuning or black-box optimization to generate a fixed draft model, relying instead on simple rules to generate varying draft models adapted to the input context. We show empirically that our light-weight algorithm is competitive with the current SOTA for self-speculative decoding, while being a truly plug-and-play method.

cs.CL↗

arXiv2024

A study on the $F$-distribution motivated by Chvátal's theorem

Let $X_{d_1, d_2}$ be an $F$-random variable with parameters $d_1$ and $d_2,$ and expectation $E[X_{d_1, d_2}]$. In this paper, for any $κ>0,$ we investigate the infimum value of the probability $P(X_{d_1, d_2}\leq κE[X_{d_1, d_2}])$. Our motivation comes from Chvátal's theorem on the binomial distribution.

math.PR↗

arXiv2024

Heterogeneous LiDAR Dataset for Benchmarking Robust Localization in Diverse Degenerate Scenarios

The ability to estimate pose and generate maps using 3D LiDAR significantly enhances robotic system autonomy. However, existing open-source datasets lack representation of geometrically degenerate environments, limiting the development and benchmarking of robust LiDAR SLAM algorithms. To address this gap, we introduce GEODE, a comprehensive multi-LiDAR, multi-scenario dataset specifically designed to include real-world geometrically degenerate environments. GEODE comprises 64 trajectories spanning over 64 kilometers across seven diverse settings with varying degrees of degeneracy. The data was meticulously collected to promote the development of versatile algorithms by incorporating various LiDAR sensors, stereo cameras, IMUs, and diverse motion conditions. We evaluate state-of-the-art SLAM approaches using the GEODE dataset to highlight current limitations in LiDAR SLAM techniques. This extensive dataset will be publicly available at https://geode.github.io, supporting further advancements in LiDAR-based SLAM.

cs.RO↗

arXiv2024

Self-supervised Photographic Image Layout Representation Learning

In the domain of image layout representation learning, the critical process of translating image layouts into succinct vector forms is increasingly significant across diverse applications, such as image retrieval, manipulation, and generation. Most approaches in this area heavily rely on costly labeled datasets and notably lack in adapting their modeling and learning methods to the specific nuances of photographic image layouts. This shortfall makes the learning process for photographic image layouts suboptimal. In our research, we directly address these challenges. We innovate by defining basic layout primitives that encapsulate various levels of layout information and by mapping these, along with their interconnections, onto a heterogeneous graph structure. This graph is meticulously engineered to capture the intricate layout information within the pixel domain explicitly. Advancing further, we introduce novel pretext tasks coupled with customized loss functions, strategically designed for effective self-supervised learning of these layout graphs. Building on this foundation, we develop an autoencoder-based network architecture skilled in compressing these heterogeneous layout graphs into precise, dimensionally-reduced layout representations. Additionally, we introduce the LODB dataset, which features a broader range of layout categories and richer semantics, serving as a comprehensive benchmark for evaluating the effectiveness of layout representation learning methods. Our extensive experimentation on this dataset demonstrates the superior performance of our approach in the realm of photographic image layout representation learning.

cs.CV↗

arXiv2024

Global Strong Solutions to the Cauchy Problem of Three-dimensional Isentropic Magnetohydrodynamics Equations with Large Initial Data

We consider the Cauchy problem to the three-dimensional isentropic compressible Magnetohydrodynamics (MHD) system with density-dependent viscosities. When the initial density is linearly equivalent to a large constant state, we prove that strong solutions exist globally in time, and there is no restriction on the size of the initial velocity and initial magnetic field.

math.AP↗

arXiv2024

Local existence of classical solutions to the 3D isentropic compressible Navier-Stokes-Poisson equations with degenerate viscosities and vacuum

We consider the isentropic compressible Navier-Stokes-Poisson equations with degenerate viscousities and vacuum in a three-dimensional torus. The local well-posedness of classical solution is established by introducing a "quasi-symmetric hyperbolic"-"degenerate elliptic" coupled structure to control the behavior of the velocity of the fluid near the vacuum and give some uniform estimates. In particular, the initial data allows vacuum in an open set and we do not need any initial compatibility conditions.

math.AP↗

arXiv2024

Improved Genetic Algorithm Based on Greedy and Simulated Annealing Ideas for Vascular Robot Ordering Strategy

This study presents a comprehensive approach for optimizing the acquisition, utilization, and maintenance of ABLVR vascular robots in healthcare settings. Medical robotics, particularly in vascular treatments, necessitates precise resource allocation and optimization due to the complex nature of robot and operator maintenance. Traditional heuristic methods, though intuitive, often fail to achieve global optimization. To address these challenges, this research introduces a novel strategy, combining mathematical modeling, a hybrid genetic algorithm, and ARIMA time series forecasting. Considering the dynamic healthcare environment, our approach includes a robust resource allocation model for robotic vessels and operators. We incorporate the unique requirements of the adaptive learning process for operators and the maintenance needs of robotic components. The hybrid genetic algorithm, integrating simulated annealing and greedy approaches, efficiently solves the optimization problem. Additionally, ARIMA time series forecasting predicts the demand for vascular robots, further enhancing the adaptability of our strategy. Experimental results demonstrate the superiority of our approach in terms of optimization, transparency, and convergence speed from other state-of-the-art methods.

cs.NE↗

arXiv2024

SEGAN: semi-supervised learning approach for missing data imputation

In many practical real-world applications, data missing is a very common phenomenon, making the development of data-driven artificial intelligence theory and technology increasingly difficult. Data completion is an important method for missing data preprocessing. Most existing miss-ing data completion models directly use the known information in the missing data set but ignore the impact of the data label information contained in the data set on the missing data completion model. To this end, this paper proposes a missing data completion model SEGAN based on semi-supervised learning, which mainly includes three important modules: generator, discriminator and classifier. In the SEGAN model, the classifier enables the generator to make more full use of known data and its label information when predicting missing data values. In addition, the SE-GAN model introduces a missing hint matrix to allow the discriminator to more effectively distinguish between known data and data filled by the generator. This paper theoretically proves that the SEGAN model that introduces a classifier and a missing hint matrix can learn the real known data distribution characteristics when reaching Nash equilibrium. Finally, a large number of experiments were conducted in this article, and the experimental results show that com-pared with the current state-of-the-art multivariate data completion method, the performance of the SEGAN model is improved by more than 3%.

cs.LG↗

arXiv2024

Resonance RoPE: Improving Context Length Generalization of Large Language Models

This paper addresses the challenge of train-short-test-long (TSTL) scenarios in Large Language Models (LLMs) equipped with Rotary Position Embedding (RoPE), where models pre-trained on shorter sequences face difficulty with out-of-distribution (OOD) token positions in longer sequences. We introduce Resonance RoPE, a novel approach designed to narrow the generalization gap in TSTL scenarios by refining the interpolation of RoPE features for OOD positions, significantly improving the model performance without additional online computational costs. Furthermore, we present PosGen, a new synthetic benchmark specifically designed for fine-grained behavior analysis in TSTL scenarios, aiming to isolate the constantly increasing difficulty of token generation on long contexts from the challenges of recognizing new token positions. Our experiments on synthetic tasks show that after applying Resonance RoPE, Transformers recognize OOD position better and more robustly. Our extensive LLM experiments also show superior performance after applying Resonance RoPE to the current state-of-the-art RoPE scaling method, YaRN, on both upstream language modeling tasks and a variety of downstream long-text applications.

cs.CL↗

arXiv2024

EVI-SAM: Robust, Real-time, Tightly-coupled Event-Visual-Inertial State Estimation and 3D Dense Mapping

Event cameras are bio-inspired, motion-activated sensors that demonstrate substantial potential in handling challenging situations, such as motion blur and high-dynamic range. In this paper, we proposed EVI-SAM to tackle the problem of 6 DoF pose tracking and 3D reconstruction using monocular event camera. A novel event-based hybrid tracking framework is designed to estimate the pose, leveraging the robustness of feature matching and the precision of direct alignment. Specifically, we develop an event-based 2D-2D alignment to construct the photometric constraint, and tightly integrate it with the event-based reprojection constraint. The mapping module recovers the dense and colorful depth of the scene through the image-guided event-based mapping method. Subsequently, the appearance, texture, and surface mesh of the 3D scene can be reconstructed by fusing the dense depth map from multiple viewpoints using truncated signed distance function (TSDF) fusion. To the best of our knowledge, this is the first non-learning work to realize event-based dense mapping. Numerical evaluations are performed on both publicly available and self-collected datasets, which qualitatively and quantitatively demonstrate the superior performance of our method. Our EVI-SAM effectively balances accuracy and robustness while maintaining computational efficiency, showcasing superior pose tracking and dense mapping performance in challenging scenarios. Video Demo: https://youtu.be/Nn40U4e5Si8.

cs.CV↗

arXiv2024

Knowledge Gaps and Research Needs for Modeling CO2 Mineralization in the Basalt-CO2-Water System: A Review of Laboratory Experiments

Carbon capture and storage in basalt is being actively investigated as a scalable climate change mitigation option. Accurate geochemical modeling prediction of the extent and rate of CO2 mineralization is a critical component in assessing the local and global feasibility and efficacy of this strategy. In this study, we review basalt-CO2-water interaction experimental studies conducted during the last two decades to determine whether they provide useable information for geochemical modeling. Most of the cited experiments generate data on the temporal evolution of water composition, and a few provide identification of secondary precipitates and their compositions, offering empirical and semi-quantitative information about the reactivity of basalts and the likelihood of secondary carbonate mineralization at various temperatures, pHs, and pCO2 conditions. However, most experiments provide insufficient information on the properties and quantity of secondary minerals formed, prohibiting accurate mass balance calculations and hence more quantitative geochemical modeling studies. Primary Ca, Mg, and Fe-bearing minerals in basalt control the availability of major ions released into aqueous solution for carbonate precipitation, and many secondary minerals, i.e., smectites, Ca-Mg-Fe carbonates, and zeolites, provide sinks for the same major ions, some of which are difficult to quantify experimentally. Thus, we have a multi-source and multi-sink inverse mass balance problem with insufficient constraints on the bulk system in which the temporal evolution of major ions does not provide sufficient information on which mineral(s) dissolve or the sequence of dissolution and precipitation reactions. Going forward, we propose that future experimental work should focus on trace elements and multiple isotopic tracers and better characterize the solid reaction products with modern analytical instruments.

physics.geo-ph↗

arXiv2024

RTMO: Towards High-Performance One-Stage Real-Time Multi-Person Pose Estimation

Real-time multi-person pose estimation presents significant challenges in balancing speed and precision. While two-stage top-down methods slow down as the number of people in the image increases, existing one-stage methods often fail to simultaneously deliver high accuracy and real-time performance. This paper introduces RTMO, a one-stage pose estimation framework that seamlessly integrates coordinate classification by representing keypoints using dual 1-D heatmaps within the YOLO architecture, achieving accuracy comparable to top-down methods while maintaining high speed. We propose a dynamic coordinate classifier and a tailored loss function for heatmap learning, specifically designed to address the incompatibilities between coordinate classification and dense prediction models. RTMO outperforms state-of-the-art one-stage pose estimators, achieving 1.1% higher AP on COCO while operating about 9 times faster with the same backbone. Our largest model, RTMO-l, attains 74.8% AP on COCO val2017 and 141 FPS on a single V100 GPU, demonstrating its efficiency and accuracy. The code and models are available at https://github.com/open-mmlab/mmpose/tree/main/projects/rtmo.

cs.CV↗

arXiv2024

Local Feature Matching Using Deep Learning: A Survey

Local feature matching enjoys wide-ranging applications in the realm of computer vision, encompassing domains such as image retrieval, 3D reconstruction, and object recognition. However, challenges persist in improving the accuracy and robustness of matching due to factors like viewpoint and lighting variations. In recent years, the introduction of deep learning models has sparked widespread exploration into local feature matching techniques. The objective of this endeavor is to furnish a comprehensive overview of local feature matching methods. These methods are categorized into two key segments based on the presence of detectors. The Detector-based category encompasses models inclusive of Detect-then-Describe, Joint Detection and Description, Describe-then-Detect, as well as Graph Based techniques. In contrast, the Detector-free category comprises CNN Based, Transformer Based, and Patch Based methods. Our study extends beyond methodological analysis, incorporating evaluations of prevalent datasets and metrics to facilitate a quantitative comparison of state-of-the-art techniques. The paper also explores the practical application of local feature matching in diverse domains such as Structure from Motion, Remote Sensing Image Registration, and Medical Image Registration, underscoring its versatility and significance across various fields. Ultimately, we endeavor to outline the current challenges faced in this domain and furnish future research directions, thereby serving as a reference for researchers involved in local feature matching and its interconnected domains. A comprehensive list of studies in this survey is available at https://github.com/vignywang/Awesome-Local-Feature-Matching .

cs.CV↗

arXiv2024

ESVIO: Event-based Stereo Visual Inertial Odometry

Event cameras that asynchronously output low-latency event streams provide great opportunities for state estimation under challenging situations. Despite event-based visual odometry having been extensively studied in recent years, most of them are based on monocular and few research on stereo event vision. In this paper, we present ESVIO, the first event-based stereo visual-inertial odometry, which leverages the complementary advantages of event streams, standard images and inertial measurements. Our proposed pipeline achieves temporal tracking and instantaneous matching between consecutive stereo event streams, thereby obtaining robust state estimation. In addition, the motion compensation method is designed to emphasize the edge of scenes by warping each event to reference moments with IMU and ESVIO back-end. We validate that both ESIO (purely event-based) and ESVIO (event with image-aided) have superior performance compared with other image-based and event-based baseline methods on public and self-collected datasets. Furthermore, we use our pipeline to perform onboard quadrotor flights under low-light environments. A real-world large-scale experiment is also conducted to demonstrate long-term effectiveness. We highlight that this work is a real-time, accurate system that is aimed at robust state estimation under challenging environments.

cs.RO↗

arXiv2024

Hyperparameter Optimization for Large Language Model Instruction-Tuning

The fine-tuning of Large Language Models (LLMs) has enabled them to recently achieve milestones in natural language processing applications. The emergence of ever larger LLMs has paved the way for more efficient fine-tuning methods. Among these, the Low-Rank Adaptation (LoRA) method keeps most of the weights of the pre-trained LLM frozen while introducing a low-rank decomposition of the weight matrix, enabling the tuning of only a very small proportion of the network. The performance on downstream tasks of models fine-tuned with LoRA heavily relies on a set of hyperparameters including the rank of the decomposition. In this work, we investigate the choice of these hyperparameters through two main blackbox optimization (BBO) techniques. We examine the whole pipeline of performing fine-tuning and validation on a pre-trained LLM as a blackbox and efficiently explore the space of hyperparameters with the \nomad algorithm, achieving a boost in performance and human alignment of the tuned model.

cs.CL↗

arXiv2024

The infimum values of the probability functions for some infinitely divisible distributions motivated by Chvátal's theorem

Let $B(n,p)$ denote a binomial random variable with parameters $n$ and $p$. Chvátal's theorem says that for any fixed $n\geq 2$, as $m$ ranges over $\{0,\ldots,n\}$, the probability $q_m:=P(B(n,m/n)\leq m)$ is the smallest when $m$ is closest to $\frac{2n}{3}$. Motivated by this theorem, in this paper we consider the infimum value of the probability $P(X\leq κE[X])$, where $κ$ is a positive real number, and $X$ is a random variable whose distribution belongs to some infinitely divisible distributions including the inverse Gaussian, log-normal, Gumbel and logistic distributions.

math.PR↗

arXiv2023

FAPP: Fast and Adaptive Perception and Planning for UAVs in Dynamic Cluttered Environments

Obstacle avoidance for Unmanned Aerial Vehicles (UAVs) in cluttered environments is significantly challenging. Existing obstacle avoidance for UAVs either focuses on fully static environments or static environments with only a few dynamic objects. In this paper, we take the initiative to consider the obstacle avoidance of UAVs in dynamic cluttered environments in which dynamic objects are the dominant objects. This type of environment poses significant challenges to both perception and planning. Multiple dynamic objects possess various motions, making it extremely difficult to estimate and predict their motions using one motion model. The planning must be highly efficient to avoid cluttered dynamic objects. This paper proposes Fast and Adaptive Perception and Planning (FAPP) for UAVs flying in complex dynamic cluttered environments. A novel and efficient point cloud segmentation strategy is proposed to distinguish static and dynamic objects. To address multiple dynamic objects with different motions, an adaptive estimation method with covariance adaptation is proposed to quickly and accurately predict their motions. Our proposed trajectory optimization algorithm is highly efficient, enabling it to avoid fast objects. Furthermore, an adaptive re-planning method is proposed to address the case when the trajectory optimization cannot find a feasible solution, which is common for dynamic cluttered environments. Extensive validations in both simulation and real-world experiments demonstrate the effectiveness of our proposed system for highly dynamic and cluttered environments.

cs.RO↗

arXiv2023

ECMD: An Event-Centric Multisensory Driving Dataset for SLAM

Leveraging multiple sensors enhances complex environmental perception and increases resilience to varying luminance conditions and high-speed motion patterns, achieving precise localization and mapping. This paper proposes, ECMD, an event-centric multisensory dataset containing 81 sequences and covering over 200 km of various challenging driving scenarios including high-speed motion, repetitive scenarios, dynamic objects, etc. ECMD provides data from two sets of stereo event cameras with different resolutions (640*480, 346*260), stereo industrial cameras, an infrared camera, a top-installed mechanical LiDAR with two slanted LiDARs, two consumer-level GNSS receivers, and an onboard IMU. Meanwhile, the ground-truth of the vehicle was obtained using a centimeter-level high-accuracy GNSS-RTK/INS navigation system. All sensors are well-calibrated and temporally synchronized at the hardware level, with recording data simultaneously. We additionally evaluate several state-of-the-art SLAM algorithms for benchmarking visual and LiDAR SLAM and identifying their limitations. The dataset is available at https://arclab-hku.github.io/ecmd/.

cs.RO↗