arXiv subjects

Peng Yin

Publications and source records attributed to Peng Yin.

At least 19 recordsLinked to original sources

arXiv2026

EgoNav: Bridging Learned Waypoints and Geometry-Aware Local Control for Robust Indoor Navigation

Image-goal navigation using lightweight topological maps is a practical paradigm for indoor robot deployment: the map requires only geotagged images, and localization relies on visual matching rather than precise pose estimation. However, learned waypoint predictors can produce targets that violate geometric constraints or deviate from the global path. Executing these waypoints safely further requires a local planner capable of collision avoidance, yet existing systems either lack one or rely on fixed parameters that cannot adapt to confined spaces. To address these limitations while retaining the navigational intuition of the learned predictor, we present EgoNav, a hierarchical system that implements this idea by generating candidates from semantically segmented traversable regions and scoring them alongside the learned waypoint for geometric safety, directional coherence, and fidelity to the learned prior. An adaptive local path planner then executes the refined waypoint with parameters modulated based on the refinement outcome. Experiments in Habitat-sim and on a physical humanoid robot show that EgoNav consistently outperforms contemporary baselines in both success rate and path efficiency.

cs.RO↗

arXiv2026

EvoPINN: Agentic Discovery of Executable Algorithms for Physics-Informed Neural Networks

Physics-informed neural networks (PINNs) have emerged as a powerful paradigm for solving partial differential equations (PDEs), yet their performance heavily relies on the manual, trial-and-error engineering of neural representations, loss formulations, and optimization dynamics. While Large Language Models (LLMs) offer a promising avenue for automated design, unconstrained code generation often yields mathematically invalid or numerically unstable solutions under strict scientific computing constraints. To bridge this gap, we propose \textbf{EvoPINN}, an agentic framework that reformulates PINN development from labor-intensive manual design into a rigorous, execution-grounded algorithm discovery problem. EvoPINN navigates a modular search space by decoupling neural representations from training programs, utilizing an LLM agent to iteratively propose memory-conditioned programmatic modifications. To ensure scientific validity, all candidates undergo strict structural verification and budget-matched PDE evaluation. Extensive experiments across diverse PDE regimes (oscillatory, elliptic, dissipative, and nonlinear transport) demonstrate that EvoPINN discovers PDE-specialized learning algorithms that significantly reduce relative $L_{2}$ error compared to baselines. Crucially, EvoPINN autonomously invented SLRC-PINN, a novel architecture whose performance gains persist under rigorous parameter-matched comparisons, establishing the viability of execution-grounded agents for discovering genuinely new scientific computing mechanisms.

cs.AI↗

arXiv2026

DCVC-MB: Neural B-Frame Video Compression using State Space Models

In this paper we propose DCVC-Mamba (DCVC-MB), a neural video codec framework for B-frame coding. Our approach incorporates an IBP frame strategy for low-delay B-frame coding, a spatio-temporal fusion model based on state-space models for bidirectional temporal prediction, and an entropy-aware skipping mechanism that selectively omits coding certain latents to reduce entropy coding times. In addition to our model contributions we also implement two inference-time strategies that enhance compression performance. Experimental evaluation shows that DCVC-MB compares favorably to existing NVCs and traditional codecs. The method demonstrates BD-rate reductions of up to $8.98\%$ on average compared to prior neural video codecs, and improvements of up to $30.45\%$ and $1.81\%$ over the VTM-19.0-LDP and VTM-19.0-RA(Inter-GoP=16) benchmarks, respectively, contributing to advances in neural video compression.

cs.CV↗

arXiv2026

MixCompress: Mixture of Experts for Variable Rate Learned Image Compression

Learned image compression (LIC) is bottlenecked by the need to store independent models for each rate-distortion operating point. Existing variable bit-rate (VBR) methods aim to reduce this overhead via dense parameter modulation, but forcing a shared backbone to approximate divergent mappings causes severe feature entanglement. Specifically, low-rate smoothing gradients inherently conflict with the preservation of high-frequency textural details, leading to sub-optimal performance. To resolve this, we propose MixCompress, a unified VBR framework based on sparse structural specialization. While sparsely gated Mixture-of-Experts (MoE) routing successfully mitigates gradient conflict, it operates on a fixed computational budget. To address the increased representational demands of higher bit-rates we introduce a Mixture-of-Depths (MoD) extension to dynamically scale model capacity. Combined with Conditional Auxiliary Transforms (CAT) for dynamic sub-band energy modulation, our hierarchical framework effectively dynamically scales capacity. Extensive evaluations demonstrate that MixCompress not only matches individually optimized single-rate baselines but can even surpass them, establishing a new Pareto frontier for computationally efficient image coding.

cs.CV↗

arXiv2025

Halo structure of $^6$He from $\textit{ab initio}$ two-nucleon spatial correlations

We evaluate pairwise correlations using ground state wave functions for $^4$He and $^6$He obtained by $\textit{ab initio}$ no-core shell model calculations with the Daejeon16 nucleon-nucleon interaction plus Coulomb interaction, to characterize the structures of these two systems. We demonstrate that two-nucleon spatial correlations, specifically the pair-number operator $r^0$ and the square-separation operator $r^2$ projected on two-body spin $S$ and isospin $z$ components encode important details of the halo structure of $^6$He. We also analyze the single-particle state occupancies and the two-body state occupancies for the ground state of $^4$He and $^6$He. Our results indicate that the two valence neutrons in the ground state of $^6$He dominantly form a spin-singlet configuration. The rms pair separations between core nucleons and halo neutrons of $^6$He are, on average, about 80% larger than pair separations within the swollen and off-centered "$\alpha$ core". We show that this off-centering effect is primarily responsible for the observed increase in point-proton radius $r_p$ in $^6$He relative to $^4$He.

nucl-th↗

arXiv2025

INT-DTT+: Low-Complexity Data-Dependent Transforms for Video Coding

Discrete trigonometric transforms (DTTs), such as the DCT-2 and the DST-7, are widely used in video codecs for their balance between coding performance and computational efficiency. In contrast, data-dependent transforms, such as the Karhunen-Lo\`eve transform (KLT) and graph-based separable transforms (GBSTs), offer better energy compaction but lack symmetries that can be exploited to reduce computational complexity. This paper bridges this gap by introducing a general framework to design low-complexity data-dependent transforms. Our approach builds on DTT+, a family of GBSTs derived from rank-one updates of the DTT graphs, which can adapt to signal statistics while retaining a structure amenable to fast computation. We first propose a graph learning algorithm for DTT+ that estimates the rank-one updates for rows and column graphs jointly, capturing the statistical properties of the overall block. Then, we exploit the progressive structure of DTT+ to decompose the kernel into a base DTT and a structured Cauchy matrix. By leveraging low-complexity integer DTTs and sparsifying the Cauchy matrix, we construct an integer approximation to DTT+, termed INT-DTT+. This approximation significantly reduces both computational and memory complexities with respect to the separable KLT with minimal performance loss. We validate our approach in the context of mode-dependent transforms for the VVC standard, following a rate-distortion optimized transform (RDOT) design approach. Integrated into the explicit multiple transform selection (MTS) framework of VVC in a rate-distortion optimization setup, INT-DTT+ achieves more than 3% BD-rate savings over the VVC MTS baseline, with complexity comparable to the integer DCT-2 once the base DTT coefficients are available.

eess.IV↗

arXiv2025

Region-Adaptive Learned Hierarchical Encoding for 3D Gaussian Splatting Data

We introduce Region-Adaptive Learned Hierarchical Encoding (RALHE) for 3D Gaussian Splatting (3DGS) data. While 3DGS has recently become popular for novel view synthesis, the size of trained models limits its deployment in bandwidth-constrained applications such as volumetric media streaming. To address this, we propose a learned hierarchical latent representation that builds upon the principles of "overfitted" learned image compression (e.g., Cool-Chic and C3) to efficiently encode 3DGS attributes. Unlike images, 3DGS data have irregular spatial distributions of Gaussians (geometry) and consist of multiple attributes (signals) defined on the irregular geometry. Our codec is designed to account for these differences between images and 3DGS. Specifically, we leverage the octree structure of the voxelized 3DGS geometry to obtain a hierarchical multi-resolution representation. Our approach overfits latents to each Gaussian attribute under a global rate constraint. These latents are decoded independently through a lightweight decoder network. To estimate the bitrate during training, we employ an autoregressive probability model that leverages octree-derived contexts from the 3D point structure. The multi-resolution latents, decoder, and autoregressive entropy coding networks are jointly optimized for each Gaussian attribute. Experiments demonstrate that the proposed RALHE compression framework achieves a rendering PSNR gain of up to 2dB at low bitrates (less than 1 MB) compared to the baseline 3DGS compression methods.

eess.IV↗

arXiv2025

Generative Distribution Embeddings: Lifting autoencoders to the space of distributions for multiscale representation learning

Many real-world problems require reasoning across multiple scales, demanding models which operate not on single data points, but on entire distributions. We introduce generative distribution embeddings (GDE), a framework that lifts autoencoders to the space of distributions. In GDEs, an encoder acts on sets of samples, and the decoder is replaced by a generator which aims to match the input distribution. This framework enables learning representations of distributions by coupling conditional generative models with encoder networks which satisfy a criterion we call distributional invariance. We show that GDEs learn predictive sufficient statistics embedded in the Wasserstein space, such that latent GDE distances approximately recover the $W_2$ distance, and latent interpolation approximately recovers optimal transport trajectories for Gaussian and Gaussian mixture distributions. We systematically benchmark GDEs against existing approaches on synthetic datasets, demonstrating consistently stronger performance. We then apply GDEs to six key problems in computational biology: learning donor-level representations from single-nuclei RNA sequencing data (6M cells), capturing clonal dynamics in lineage-traced RNA sequencing data (150K cells), predicting perturbation effects on transcriptomes (1M cells), predicting perturbation effects on cellular phenotypes (20M single-cell images), designing synthetic yeast promoters (34M sequences), and spatiotemporal modeling of viral protein sequences (1M sequences).

cs.LG↗

arXiv2025

Multi-Agent DRL for Multi-Objective Twin Migration Routing with Workload Prediction in 6G-enabled IoV

Sixth Generation (6G)-enabled Internet of Vehicles (IoV) facilitates efficient data synchronization through ultra-fast bandwidth and high-density connectivity, enabling the emergence of Vehicle Twins (VTs). As highly accurate replicas of vehicles, VTs can support intelligent vehicular applications for occupants in 6G-enabled IoV. Thanks to the full coverage capability of 6G, resource-constrained vehicles can offload VTs to edge servers, such as roadside units, unmanned aerial vehicles, and satellites, utilizing their computing and storage resources for VT construction and updates. However, communication between vehicles and edge servers with limited coverage is prone to interruptions due to the dynamic mobility of vehicles. Consequently, VTs must be migrated among edge servers to maintain uninterrupted and high-quality services for users. In this paper, we introduce a VT migration framework in 6G-enabled IoV. Specifically, we first propose a Long Short-Term Memory (LSTM)-based Transformer model to accurately predict long-term workloads of edge servers for migration decision-making. Then, we propose a Dynamic Mask Multi-Agent Proximal Policy Optimization (DM-MAPPO) algorithm to identify optimal migration routes in the highly complex environment of 6G-enabled IoV. Finally, we develop a practical platform to validate the effectiveness of the proposed scheme using real datasets. Simulation results demonstrate that the proposed DM-MAPPO algorithm significantly reduces migration latency by 20.82% and packet loss by 75.07% compared with traditional deep reinforcement learning algorithms.

cs.NI↗

arXiv2025

GelFusion: Enhancing Robotic Manipulation under Visual Constraints via Visuotactile Fusion

Visuotactile sensing offers rich contact information that can help mitigate performance bottlenecks in imitation learning, particularly under vision-limited conditions, such as ambiguous visual cues or occlusions. Effectively fusing visual and visuotactile modalities, however, presents ongoing challenges. We introduce GelFusion, a framework designed to enhance policies by integrating visuotactile feedback, specifically from high-resolution GelSight sensors. GelFusion using a vision-dominated cross-attention fusion mechanism incorporates visuotactile information into policy learning. To better provide rich contact information, the framework's core component is our dual-channel visuotactile feature representation, simultaneously leveraging both texture-geometric and dynamic interaction features. We evaluated GelFusion on three contact-rich tasks: surface wiping, peg insertion, and fragile object pick-and-place. Outperforming baselines, GelFusion shows the value of its structure in improving the success rate of policy learning.

cs.RO↗

arXiv2025

A measurement noise scaling law for cellular representation learning

Large genomic and imaging datasets can be used to train models that learn meaningful representations of cellular systems. Across domains, model performance improves predictably with dataset size and compute budget, providing a basis for allocating data and computation. Scientific data, however, is also limited by noise arising from factors such as molecular undersampling, sequencing errors, and image resolution. By fitting 1,670 representation learning models across three data modalities (gene expression, sequence, and image data), we show that noise defines a distinct axis along which performance improves. Noise scaling follows a logarithmic law. We derive the law from a model of noise propagation, and use it to define noise sensitivity and model capacity as benchmarking metrics. We show that protein sequence representations are noise-robust while single cell transcriptomics models are not, with a Transformer-based model showing greater noise robustness but lower saturating performance than a variational autoencoder model. Noise scaling metrics may support future model evaluation and experimental design.

q-bio.QM↗

arXiv2024

Standardizing Generative Face Video Compression using Supplemental Enhancement Information

This paper proposes a Generative Face Video Compression (GFVC) approach using Supplemental Enhancement Information (SEI), where a series of compact spatial and temporal representations of a face video signal (e.g., 2D/3D keypoints, facial semantics and compact features) can be coded using SEI messages and inserted into the coded video bitstream. At the time of writing, the proposed GFVC approach using SEI messages has been included into a draft amendment of the Versatile Supplemental Enhancement Information (VSEI) standard by the Joint Video Experts Team (JVET) of ISO/IEC JTC 1/SC 29 and ITU-T SG21, which will be standardized as a new version of ITU-T H.274 | ISO/IEC 23002-7. To the best of the authors' knowledge, the JVET work on the proposed SEI-based GFVC approach is the first standardization activity for generative video compression. The proposed SEI approach has not only advanced the reconstruction quality of early-day Model-Based Coding (MBC) via the state-of-the-art generative technique, but also established a new SEI definition for future GFVC applications and deployment. Experimental results illustrate that the proposed SEI-based GFVC approach can achieve remarkable rate-distortion performance compared with the latest Versatile Video Coding (VVC) standard, whilst also potentially enabling a wide variety of functionalities including user-specified animation/filtering and metaverse-related applications.

cs.CV↗

arXiv2024

Approximating mutual information of high-dimensional variables using learned representations

Mutual information (MI) is a general measure of statistical dependence with widespread application across the sciences. However, estimating MI between multi-dimensional variables is challenging because the number of samples necessary to converge to an accurate estimate scales unfavorably with dimensionality. In practice, existing techniques can reliably estimate MI in up to tens of dimensions, but fail in higher dimensions, where sufficient sample sizes are infeasible. Here, we explore the idea that underlying low-dimensional structure in high-dimensional data can be exploited to faithfully approximate MI in high-dimensional settings with realistic sample sizes. We develop a method that we call latent MI (LMI) approximation, which applies a nonparametric MI estimator to low-dimensional representations learned by a simple, theoretically-motivated model architecture. Using several benchmarks, we show that unlike existing techniques, LMI can approximate MI well for variables with $> 10^3$ dimensions if their dependence structure has low intrinsic dimensionality. Finally, we showcase LMI on two open problems in biology. First, we approximate MI between protein language model (pLM) representations of interacting proteins, and find that pLMs encode non-trivial information about protein-protein interactions. Second, we quantify cell fate information contained in single-cell RNA-seq (scRNA-seq) measurements of hematopoietic stem cells, and find a sharp transition during neutrophil differentiation when fate information captured by scRNA-seq increases dramatically.

q-bio.QM↗

arXiv2024

LF-3PM: a LiDAR-based Framework for Perception-aware Planning with Perturbation-induced Metric

Just as humans can become disoriented in featureless deserts or thick fogs, not all environments are conducive to the Localization Accuracy and Stability (LAS) of autonomous robots. This paper introduces an efficient framework designed to enhance LiDAR-based LAS through strategic trajectory generation, known as Perception-aware Planning. Unlike vision-based frameworks, the LiDAR-based requires different considerations due to unique sensor attributes. Our approach focuses on two main aspects: firstly, assessing the impact of LiDAR observations on LAS. We introduce a perturbation-induced metric to provide a comprehensive and reliable evaluation of LiDAR observations. Secondly, we aim to improve motion planning efficiency. By creating a Static Observation Loss Map (SOLM) as an intermediary, we logically separate the time-intensive evaluation and motion planning phases, significantly boosting the planning process. In the experimental section, we demonstrate the effectiveness of the proposed metrics across various scenes and the feature of trajectories guided by different metrics. Ultimately, our framework is tested in a real-world scenario, enabling the robot to actively choose topologies and orientations preferable for localization. The source code is accessible at https://github.com/ZJU-FAST-Lab/LF-3PM.

cs.RO↗

arXiv2024

Direct ab initio calculation of the $^{4}$He nuclear electric dipole polarizability

The calculation of nuclear electromagnetic sum rules by directly diagonalizing the nuclear Hamiltonian in a large basis is numerically challenging and has not been performed for $A>2$ nuclei. With the significant progress of high performance computing, we show that calculating sum rules using numerous discretized continuum states obtained by directly diagonalizing the ab initio no-core shell model Hamiltonian is achievable numerically. Specifically, we calculate the $^{4}$He electric dipole ($E1$) polarizability, that is an inverse energy weighted sum rule, employing the Daejeon16 $NN$ interaction. We demonstrate that the calculations are numerically tractable as the dimension of the basis increases and are convergent. Our results for the $^{4}$He electric dipole polarizability are consistent with the most recent experimental data and are compared with those of other theoretical studies employing different techniques and various interactions.

nucl-th↗

arXiv2024

General Place Recognition Survey: Towards Real-World Autonomy

In the realm of robotics, the quest for achieving real-world autonomy, capable of executing large-scale and long-term operations, has positioned place recognition (PR) as a cornerstone technology. Despite the PR community's remarkable strides over the past two decades, garnering attention from fields like computer vision and robotics, the development of PR methods that sufficiently support real-world robotic systems remains a challenge. This paper aims to bridge this gap by highlighting the crucial role of PR within the framework of Simultaneous Localization and Mapping (SLAM) 2.0. This new phase in robotic navigation calls for scalable, adaptable, and efficient PR solutions by integrating advanced artificial intelligence (AI) technologies. For this goal, we provide a comprehensive review of the current state-of-the-art (SOTA) advancements in PR, alongside the remaining challenges, and underscore its broad applications in robotics. This paper begins with an exploration of PR's formulation and key research challenges. We extensively review literature, focusing on related methods on place representation and solutions to various PR challenges. Applications showcasing PR's potential in robotics, key PR datasets, and open-source libraries are discussed. We conclude with a discussion on PR's future directions and provide a summary of the literature covered at: https://github.com/MetaSLAM/GPRS.

cs.RO↗

arXiv2023

Experimental super-Heisenberg quantum metrology with indefinite gate order

The precision of quantum metrology is widely believed to be restricted by the Heisenberg limit, corresponding to a root mean square error that is inversely proportional to the number of independent processes probed in an experiment, N. In the past, some proposals have challenged this belief, for example using non-linear interactions among the probes. However, these proposals turned out to still obey the Heisenberg limit with respect to other relevant resources, such as the total energy of the probes. Here, we present a photonic implementation of a quantum metrology protocol surpassing the Heisenberg limit by probing two groups of independent processes in a superposition of two alternative causal orders. Each process creates a phase space displacement, and our setup is able to estimate a geometric phase associated to two sets of N displacements with an error that falls quadratically with N. Our results only require a single-photon probe with an initial energy that is independent of N. Using a superposition of causal orders outperforms every setup where the displacements are probed in a definite order. Our experiment features the demonstration of indefinite causal order in a continuous-variable system, and opens up the experimental investigation of quantum metrology setups boosted by indefinite causal order.

quant-ph↗

arXiv2022

A local-density-approximation description of high-momentum tails in isospin asymmetric nuclei

We adapt the local density approximation to add the high-momentum tails (HMTs) to finite nuclei's Slater-determinant momentum distributions. The HMTs are extracted by the extended Brueckner-Hartree-Fock (EBHF) method or by the lowest order cluster approximation. With a correction factor being added to EBHF, it is sufficiently in agreement with the experimental benchmark, i.e., the high-momentum $N/Z$ ratios approximately equal to $1$, and the low-momentum $N/Z$ ratios approximately equal to $N/Z$ of the systems. It is also found that the tensor force makes the nucleon-nucleon correlations appear more easily on the nuclear surface region and the percentage of high-momentum ($p > 300$ MeV/c) nucleons, around $17\%$--$18\%$, independent of isospin asymmetry.

nucl-th↗