arXiv subjects

Yiming Yan

Publications and source records attributed to Yiming Yan.

12 recordsLinked to original sources

arXiv2026

GASE: Gaussian Splatting-Based Automated System for Reconstructing Embodied-Simulation Environments

Training embodied agents in the real world requires skilled operators and expensive hardware. Simulation environments offer a compelling alternative by enabling large-scale, cost-effective data augmentation. Consequently, rapidly constructing high-fidelity simulation scenes with a minimal sim-to-real gap has become a critical objective in robot learning. While reconstruction-based methods provide superior visual quality, current workflows are hindered by inefficient data acquisition and subpar foreground object extraction. We thus propose GASE, a highly automated system for simulation scene construction. GASE leverages multi-view video streams from panoramic camera arrays to enable rapid environment scanning. To ensure high-quality asset generation, our pipeline introduces a camera-pose-based strategy that robustly extracts objects across frames in the 2D domain, followed by high-fidelity scene inpainting. Foreground objects and the static background are then reconstructed independently and seamlessly imported into physics simulators for policy training. Extensive experiments demonstrate that GASE outperforms existing 3D Gaussian-based methods in segmentation accuracy by over 10\% while achieving state-of-the-art inpainting quality. Furthermore, real-robot deployments across manipulation and navigation tasks maintains a performance gap of less than 10\% compared to policies trained purely on real-world data. These results confirm that GASE provides an efficient and highly effective solution for bridging the sim-to-real gap. Code will be released.

cs.RO↗

arXiv2026

LingxiDiagBench: A Multi-Agent Framework for Benchmarking LLMs in Chinese Psychiatric Consultation and Diagnosis

Mental disorders are highly prevalent worldwide, but the shortage of psychiatrists and the inherent subjectivity of interview-based diagnosis create substantial barriers to timely and consistent mental-health assessment. Progress in AI-assisted psychiatric diagnosis is constrained by the absence of benchmarks that simultaneously provide realistic patient simulation, clinician-verified diagnostic labels, and support for dynamic multi-turn consultation. We present LingxiDiagBench, a large-scale multi-agent benchmark that evaluates LLMs on both static diagnostic inference and dynamic multi-turn psychiatric consultation in Chinese. At its core is LingxiDiag-16K, a dataset of 16,000 EMR-aligned synthetic consultation dialogues designed to reproduce real clinical demographic and diagnostic distributions across 12 ICD-10 psychiatric categories. Through extensive experiments across state-of-the-art LLMs, we establish key findings: (1) although LLMs achieve high accuracy on binary depression--anxiety classification (up to 92.3%), performance deteriorates substantially for depression--anxiety comorbidity recognition (43.0%) and 12-way differential diagnosis (28.5%); (2) dynamic consultation often underperforms static evaluation, indicating that ineffective information-gathering strategies significantly impair downstream diagnostic reasoning; (3) consultation quality assessed by LLM-as-a-Judge shows only moderate correlation with diagnostic accuracy, suggesting that well-structured questioning alone does not ensure correct diagnostic decisions. We release LingxiDiag-16K and the full evaluation framework to support reproducible research at https://github.com/Lingxi-mental-health/LingxiDiagBench.

cs.AI↗

arXiv2025

Asking like Socrates: Socrates helps VLMs understand remote sensing images

Recent multimodal reasoning models, inspired by DeepSeek-R1, have significantly advanced vision-language systems. However, in remote sensing (RS) tasks, we observe widespread pseudo reasoning: models narrate the process of reasoning rather than genuinely reason toward the correct answer based on visual evidence. We attribute this to the Glance Effect, where a single, coarse perception of large-scale RS imagery results in incomplete understanding and reasoning based on linguistic self-consistency instead of visual evidence. To address this, we propose RS-EoT (Remote Sensing Evidence-of-Thought), a language-driven, iterative visual evidence-seeking paradigm. To instill this paradigm, we propose SocraticAgent, a self-play multi-agent system that synthesizes reasoning traces via alternating cycles of reasoning and visual inspection. To enhance and generalize these patterns, we propose a two-stage progressive RL strategy: first, RL on fine-grained Grounding tasks to enhance RS-EoT capabilities, followed by RL on RS VQA to generalize to broader understanding scenarios. Experiments show RS-EoT achieves state-of-the-art performance on multiple RS VQA and grounding benchmarks. Analyses reveal clear iterative cycles of reasoning and evidence seeking, confirming RS-EoT mitigates the Glance Effect and enables genuine evidence-grounded reasoning. Our code, data, and models are available at https://geox-lab.github.io/Asking_like_Socrates

cs.CV↗

arXiv2025

Reality-Infused Deep Learning for Angle-resolved Quasi-optical Fourier Surfaces

Optical Fourier surfaces (OFSs), featuring sinusoidally profiled diffractive elements, manipulate light through patterned nanostructures and incident angle modulation. Compared to altering structural parameters, tuning elevation and azimuth angles offers greater design flexibility for light field control. However, angle-resolved responses of OFSs are often complex due to diverse mode excitations and couplings, complicating the alignment between simulations and practical fabrication. Here, we present a reality-infused deep learning framework, empowered by angle-resolved measurements, to enable real-time and accurate predictions of angular dispersion in quasi-OFSs. This approach captures critical features, including nanofabrication and measurement imperfections, which conventional simulation-based methods typically overlook. Our framework significantly accelerates the design process while achieving predictive performance highly consistent with experimental observations across broad angular and spectral ranges. Our study supports valuable insights into the development of OFS-based devices, and represents a paradigm shift from simulation-driven to reality-infused methods, paving the way for advancements in optical design applications.

physics.optics↗

arXiv2023

State-Dependent Channels with a Message-Cognizant Helper

The capacity of a state-dependent discrete memoryless channel (SD-DMC) is derived for the setting where a message-cognizant rate-limited helper observes the state sequence noncausally, produces its description, and provides the description to both encoder and decoder.

cs.IT↗

arXiv2023

Message-Cognizant Assistance and Feedback for the Gaussian Channel

A formula is derived for the capacity of the Gaussian channel with a benevolent message-cognizant rate-limited helper that provides a noncausal description of the noise to the encoder and decoder. This capacity is strictly larger than when the helper is message oblivious, with the difference being particularly pronounced at low signal-to-noise ratios. It is shown that in this setup, a feedback link from the receiver to the encoder does not increase capacity. However, in the presence of such a link, said capacity can be achieved even if the helper is oblivious to the transmitted message.

cs.IT↗

arXiv2023

BLEURT Has Universal Translations: An Analysis of Automatic Metrics by Minimum Risk Training

Automatic metrics play a crucial role in machine translation. Despite the widespread use of n-gram-based metrics, there has been a recent surge in the development of pre-trained model-based metrics that focus on measuring sentence semantics. However, these neural metrics, while achieving higher correlations with human evaluations, are often considered to be black boxes with potential biases that are difficult to detect. In this study, we systematically analyze and compare various mainstream and cutting-edge automatic metrics from the perspective of their guidance for training machine translation systems. Through Minimum Risk Training (MRT), we find that certain metrics exhibit robustness defects, such as the presence of universal adversarial translations in BLEURT and BARTScore. In-depth analysis suggests two main causes of these robustness deficits: distribution biases in the training datasets, and the tendency of the metric paradigm. By incorporating token-level constraints, we enhance the robustness of evaluation metrics, which in turn leads to an improvement in the performance of machine translation systems. Codes are available at \url{https://github.com/powerpuffpomelo/fairseq_mrt}.

cs.CL↗

arXiv2019

Amortized Inference of Variational Bounds for Learning Noisy-OR

Classical approaches for approximate inference depend on cleverly designed variational distributions and bounds. Modern approaches employ amortized variational inference, which uses a neural network to approximate any posterior without leveraging the structures of the generative models. In this paper, we propose Amortized Conjugate Posterior (ACP), a hybrid approach taking advantages of both types of approaches. Specifically, we use the classical methods to derive specific forms of posterior distributions and then learn the variational parameters using amortized inference. We study the effectiveness of the proposed approach on the noisy-or model and compare to both the classical and the modern approaches for approximate inference and parameter learning. Our results show that the proposed method outperforms or are at par with other approaches.

cs.LG↗

arXiv2019

Multi-surface catadioptric freeform lens design for ultra-efficient off-axis road illumination

We propose a novel design methodology to tackle the multi-surface catadioptric freeform lens design for off-axis road illumination applications based on an ideal source. The lens configuration contains an analytic refractive entrance surface, an analytic total internal reflective (TIR) surface and two freeform exit surfaces. A curl-free energy equipartition is established between the source and target plane and divided to implement the composite ray mapping mechanism. Furthermore, the analytic TIR surface and refractive entrance surface are optimized for the minimal Fresnel losses and surface error based on genetic algorithm (GA). The results show a significant improvement on illuminance uniformity and ultra-high transfer efficiency compared to our proposed result in [Zhu et al., Opt. Exp. 26, A54-A65 (2018)].

physics.optics↗

arXiv2019

Automated first-order design of double-sided telecentric zoom systems based on PSO algorithms

In this paper, we propose a new approach to automatically retrieve the first-order design of double-sided telecentric zoom systems with high efficiency based on Particle Swarm Optimization (PSO) algorithms. We transform the design problem to realize a zoom lens system with fixed positions of both the front focal point and the rear focal point during zooming. We demonstrate that the proposed algorithm is much more efficient than the Monte Carlo algorithm in designing zoom systems with two fixed foci as well as a given zoom ratio. Furthermore, a compact first-order design of three-component 4X zoom system with two fixed foci is proposed to show the high efficiency and great potential of our proposed algorithm in searching proper first-order configuration for complex optical systems.

physics.optics↗

arXiv2015

Active-set prediction for interior point methods using controlled perturbations

We propose the use of controlled perturbations to address the challenging question of optimal active-set prediction for interior point methods. Namely, in the context of linear programming, we consider perturbing the inequality constraints/bounds so as to enlarge the feasible set. We show that if the perturbations are chosen appropriately, the solution of the original problem lies on or close to the central path of the perturbed problem. We also find that a primal-dual path-following algorithm applied to the perturbed problem is able to accurately predict the optimal active set of the original problem when the duality gap for the perturbed problem is not too small; furthermore, depending on problem conditioning, this prediction can happen sooner than predicting the active set for the perturbed problem or when the original one is solved. Encouraging preliminary numerical experience is reported when comparing activity prediction for the perturbed and unperturbed problem formulations.

math.OC↗

arXiv2014

Active-set prediction in quadratic programming using interior point methods and controlled perturbations

In this paper, we extend the idea of using controlled perturbations to enhance the capabilities of active-set prediction for interior point methods for convex Quadratic Programming (QP) problems. Namely, we consider perturbing the inequality constraints (by a small amount) so as to enlarge the feasible set. We show that if the perturbations are chosen judiciously, then there exists a primal-dual pair of points which is close to the optimal solution of the perturbed problems and the corresponding active and inactive sets at this point are the same as the optimal active and inactive sets at an optimal solution of the original QP problems. Additionally, we prove that the optimal tripartition of the original problems can also be predicted by solving the perturbed ones. Furthermore, encouraging preliminary numerical experience is also presented for the QP case.

math.OC↗