arXiv subjects

Qinwu Xu

Publications and source records attributed to Qinwu Xu.

9 recordsLinked to original sources

arXiv2026

Momentum-Guided Semantic Forecasting (MoFore) for Self-Supervised Video Representation Learning

Self-supervised video representation learning has recently advanced through contrastive learning, masked reconstruction, and predictive representation learning. Reconstruction-based approaches such as MAE and VideoMAE learn representations by recovering masked visual content \cite{he2022mae,tong2022videomae}, while contrastive methods such as CLIP learn semantically meaningful embedding spaces through representation alignment \cite{radford2021clip}. In this work, we introduce a Momentum-Guided Semantic Forecasting framework (MoFore) for self-supervised video representation learning. Instead of optimizing for pixel-level reconstruction or task-specific semantic alignment, the proposed method learns temporally predictive video representations by forecasting future latent embeddings from temporally distant context clips. To improve robustness across temporal scales, we further introduce randomized temporal-gap forecasting during training. The framework combines predictive latent forecasting with contrastive regularization to encourage temporal consistency while preventing representation collapse. Experiments on the UCF101 dataset demonstrate that the proposed framework learns temporally consistent and semantically meaningful video representations without using action labels during training. Quantitative analysis shows strong temporal stability and emergent category-level structure in the learned embedding space, while qualitative retrieval experiments reveal motion-aware organization across related activities. Overall, the results suggest that long-range latent forecasting provides an effective and computationally efficient approach for self-supervised video representation learning without relying on reconstruction-based objectives.

cs.CV↗

arXiv2026

Miller-Index-Based Latent Crystallographic Fracture Plane Reasoning and generation with Vision-Language Models

We study whether multimodal large language models (MLLMs) can leverage crystallographic plane indices (Miller indices) as a structured latent representation for reasoning about fracture geometry. We formulate Miller indices $z = (h,k,l)$ as a latent variable governing idealized planar fracture and evaluate two complementary capabilities: (i) latent inference, where the model maps visual observations to plane hypotheses under physically valid conditions, and (ii) latent applicability assessment, where the model determines whether such a representation is meaningful for a given fracture image. Through extensive experiments spanning synthetic data, controlled 2D--3D geometric pairs, and real-world fracture images across multiple material classes -- including ceramics, glass, metals, and concrete -- we show that MLLMs can reliably perform latent inference in idealized settings and, critically, can reject the latent representation when the underlying physics does not support it. As an exploratory extension, we further examine AI-generated fracture sequences and observe qualitatively plausible brittle-fracture progression behaviors, suggesting that multimodal generative models may encode partial implicit physical priors related to material failure dynamics. These results suggest that MLLMs can act as physics-aware reasoning systems conditioned on structured latent priors, provided that the domain of validity is explicitly modeled.

cs.LG↗

arXiv2026

Multilingual OCR-Aware Fine-Tuning and Prompt-Guided Chain-of-Thought Reasoning for Multimodal Large Language Models

Optical character recognition (OCR) and multilingual scene-text understanding remain challenging for multimodal large language models (MLLMs), particularly in real-world images containing small or degraded text, cluttered layouts, occlusion, handwriting, and complex typography. We present an OCR-aware multilingual post-training framework that improves visual-text grounding in a general-purpose MLLM without requiring an external OCR engine, OCR-extracted text, or text bounding boxes at inference time. The framework combines large-scale multilingual OCR supervision, approximately 5M additional multilingual training samples, controlled synthetic OCR generation and in-image text translation, LoRA-based supervised fine-tuning (SFT), and lightweight OCR-oriented Chain-of-Thought prompting. On a held-out real-world multilingual OCR benchmark, OCR-SFT improves OCR completeness from 71.3 to 84.6, reduces hallucination rate from 18.3\% to 5.5\%, and improves translation BLEU-1 from 52.3 to 80.2, with substantial hallucination reductions under blur and rotation. Evaluation on public benchmarks further shows gains on OCR-intensive tasks while largely preserving broader multimodal capabilities; ablations show that SFT provides the primary improvement, with prompting offering smaller complementary gains. These results demonstrate that data-centric OCR-aware post-training provides a practical and scalable approach to improving multilingual visual-text grounding in general-purpose MLLMs.

cs.CV↗

arXiv2026

Grounded Post-Training with Hard Examples for Reducing Hallucination in Multimodal Large Language Models

Hallucination remains a fundamental challenge in vision-language models (VLMs), where autoregressive generation may produce linguistically plausible yet physically inconsistent or visually ungrounded responses due to likelihood maximization under joint probabilistic modeling. We propose a stage-wise preference optimization framework for hallucination reduction through targeted multimodal data construction. Rather than directly optimizing on generic instruction-following data, our approach progressively constructs hallucination-focused preference pairs near known failure boundaries. The framework emphasizes ambiguous spatial orientation, object relationships, OCR uncertainty, and adversarial false-premise training. Hallucinated negatives are generated through minimally perturbed yet visually inconsistent alternatives, enabling Direct Preference Optimization (DPO) to better separate grounded reasoning from plausible hallucination. Experiments on open-source benchmarks and real-world multimodal evaluation scenarios demonstrate improved grounding consistency, reduced hallucination, and more informative grounded responses. Cross-model qualitative evaluation further shows that the proposed multimodal LLM DPO framework produces more visually grounded responses than several frontier proprietary VLMs, such as in ambiguous spatial reasoning and adversarial false-premise settings. The results suggest that hallucination may arise not only from limited model capacity, but also from inherent tendencies of autoregressive probabilistic generation to favor linguistically plausible continuations under weak visual grounding. Future work may explore physical consistency modeling, uncertainty-aware multimodal reasoning, and architectural alternatives beyond standard autoregressive decoding.

cs.CV↗

arXiv2026

Robust Checkpoint Selection for Multimodal LLMs via Agentic Evaluation and Stability-Aware Ranking

Selecting a final checkpoint for multimodal large language models (MLLMs) is challenging when late-stage candidates are closely matched and downstream evaluation signals are noisy. Small observed differences can be comparable to variability introduced by finite evaluation samples, LLM-based judges, and ambiguous multimodal evidence, while validation loss may not identify the checkpoint preferred by downstream evaluation. We formulate late-stage checkpoint selection as a stability-aware decision problem under evaluation uncertainty and propose a progressive framework combining pointwise filtering, listwise ranking, and pairwise refinement. Repeated evaluation-set subsampling is used to characterize ranking stability, while percentile-based aggregation accounts for lower- and upper-tail behavior. Experiments show that multimodal data evaluability is critical: quality-aware curation of OCR-heavy inputs reduces ranking flip rate from 32.5\% to 11.2\% and increases inter-run agreement from 0.61 to 0.84. We further observe divergence between validation-loss progression and downstream checkpoint preference in two independent MLLM settings. An additional public Qwen2.5-VL-7B reproduction across 11 checkpoints shows tightly clustered pointwise scores and frequently tie-dominated final pairwise comparisons, while repeated evaluation most often selects an intermediate rather than the final checkpoint. These results suggest that reliable MLLM checkpoint selection should quantify and reserve evaluation uncertainty rather than force decisions from small differences in a single metric.

cs.LG↗

arXiv2024

A Fast Graph Search Algorithm with Dynamic Optimization and Reduced Histogram for Discrimination of Binary Classification Problem

This study develops a graph search algorithm to find the optimal discrimination path for the binary classification problem. The objective function is defined as the difference of variations between the true positive (TP) and false positive (FP). It uses the depth first search (DFS) algorithm to find the top-down paths for discrimination. It proposes a dynamic optimization procedure to optimize TP at the upper levels and then reduce FP at the lower levels. To accelerate computing speed with improving accuracy, it proposes a reduced histogram algorithm with variable bin size instead of looping over all data points, to find the feature threshold of discrimination. The algorithm is applied on top of a Support Vector Machine (SVM) model for a binary classification problem on whether a person is fit or unfit. It significantly improves TP and reduces FP of the SVM results (e.g., reduced FP by 90% with a loss of only\ 5% TP). The graph search auto-generates 39 ranked discrimination paths within 9 seconds on an input of total 328,464 objects, using a dual-core Laptop computer with a processor of 2.59 GHz.

cs.LG↗

arXiv2024

A Multi-objective Newton Optimization Algorithm for Hyper-Parameter Search

This study proposes a Newton based multiple objective optimization algorithm for hyperparameter search. The first order differential (gradient) is calculated using finite difference method and a gradient matrix with vectorization is formed for fast computation. The Newton Raphson iterative solution is used to update model parameters with iterations, and a regularization term is included to eliminate the singularity issue. The algorithm is applied to search the optimal probability threshold (a vector of eight parameters) for a multiclass object detection problem of a convolutional neural network. The algorithm quickly finds the improved parameter values to produce an overall higher true positive (TP) and lower false positive (FP) rates, as compared to using the default value of 0.5. In comparison, the Bayesian optimization generates lower performance in the testing case. However, the performance and parameter values may oscillate for some cases during iterations, which may be due to the data driven stochastic nature of the subject. Therefore, the optimal parameter value can be identified from a list of iteration steps according to the optimal TP and FP results.

math.OC↗

arXiv2021

Modeling 3D geometry using 1D laser distance measurements with application to cylinder for visualization and evaluating surface quality

Geometric metrology includes one or two-dimensional (1D or 2D) distance or plane measurements, as well as the three-dimensional (3D) scanning. The 1D or 2D measuring system is unable to obtain advanced 3D feature, while the 3D scanning system is relatively costly and time-consuming. Accordingly, in this study I developed a 3D geometry and surface prediction method by using 1D laser distance measurements to achieve 3D features while saving cost and time. The model is based on the natural neighbor function for data interpolation and linear model for extrapolation. I implemented the model to the cylinder body for evaluating 3D circularity and surface quality. Results show that the model could achieve reasonable accuracy in constructing the 3D geometry and surface deviation using limited distance measurement (e.g. only 100 data points). It accurately predicts the shape of a curved dent (30.48 mm) and identified minor dents (less than 0.18mm in depth) which are unable to be detected by eyes and finger touch. It also detects the surface corrugation (1.59 mm) and small local features (6.35 mm) using a measurement resolution of 21.34 mm by 15.49 mm. The predicted 3D circularity is higher than the measured 2D circularity as expected. The 3D model may be extended to other continuous geometry shapes for the future study.

cs.GR↗

arXiv2016

A mathematical and Physical Model Improves Accuracy in Simulating Solid Material Relaxation Modulus and Viscoelastic Responses

We propose a new material viscoelastic model and mathematical solution to simulate relaxation modulus and viscoelastic response. The model formula of relaxation modulus is extended from sigmoidal function considering nonlinear strain hardening and softening. Its physical mechanism can be interpreted by a spring network viscous medium model with only five parameters in a simpler format than the molecular-chain based polymer models to represent general materials. We also developed a three-dimensional finite-element method and robust numerical algorithms to implement this model for solving partial differential equations. We validate the model through both experimental data and numerical simulations on a broad range of materials including bitumen, shape-memory polymer, spider-inspired silk, hydrogel, biomaterials and bone. By satisfying the 2nd law of thermodynamics in the form of Calusius-Duhem inequality, the model is able to simulate creep and sinusoidal deformation, and energy dissipation. As compared to Prony series, the most general model being used often with a large number of model parameters, the proposed model has improved accuracy in fitting experimental data and predicting modulus outside of the experimental range, and the latter one is especially useful for material design. The new model also has higher numerical accuracy while competitive numerical stability and computation speed for convergence.

cond-mat.mtrl-sci↗