arXiv subjects

Mingming Zheng

Publications and source records attributed to Mingming Zheng.

3 recordsLinked to original sources

arXiv2026

Advancing Branch-and-Price for Graph Coloring: New Pricing Strategies and Benchmark Results

This paper proposes BPCOL+, an exact branch-and-price algorithm for the Graph Coloring Problem. The algorithm integrates enhanced pricing strategies based on Zero-Suppressed Binary Decision Diagrams (ZDDs) for solving the pricing problem associated with the maximal-stable-set-based set-covering formulation. After computing upper and lower bounds at the root node using heuristic procedures and column generation, BPCOL+ reduces the size of the ZDD by applying maximal stable set reduction techniques that exploit alternative dual vectors. Computational experiments on the 137 DIMACS benchmark instances and on 5,000 recently proposed Erd\H{o}s--R\'enyi instances show that BPCOL+ outperforms existing exact branch-and-price algorithms and remains highly competitive with state-of-the-art SAT-based exact solvers. In particular, BPCOL+ solves 96 DIMACS instances within one hour and proves optimality for 4,641 of the 5,000 Erd\H{o}s--R\'enyi instances.

math.OC↗

arXiv2026

Qianfan-OCR: A Unified End-to-End Model for Document Intelligence

We present Qianfan-OCR, a 4B-parameter end-to-end vision-language model that unifies document parsing, layout analysis, and document understanding within a single architecture. It performs direct image-to-Markdown conversion and supports diverse prompt-driven tasks including table extraction, chart understanding, document QA, and key information extraction. To address the loss of explicit layout analysis in end-to-end OCR, we propose Layout-as-Thought, an optional thinking phase triggered by special think tokens that generates structured layout representations -- bounding boxes, element types, and reading order -- before producing final outputs, recovering layout grounding capabilities while improving accuracy on complex layouts. Qianfan-OCR ranks first among end-to-end models on OmniDocBench v1.5 (93.12) and OlmOCR Bench (79.8), achieves competitive results on OCRBench, CCOCR, DocVQA, and ChartQA against general VLMs of comparable scale, and attains the highest average score on public key information extraction benchmarks, surpassing Gemini-3.1-Pro, Seed-2.0, and Qwen3-VL-235B. The model is publicly accessible via the Baidu AI Cloud Qianfan platform.

cs.CV↗

arXiv2025

Qianfan-VL: Domain-Enhanced Universal Vision-Language Models

We present Qianfan-VL, a series of multimodal large language models ranging from 3B to 70B parameters, achieving state-of-the-art performance through innovative domain enhancement techniques. Our approach employs multi-stage progressive training and high-precision data synthesis pipelines, which prove to be critical technologies for enhancing domain-specific capabilities while maintaining strong general performance. Qianfan-VL achieves comparable results to leading open-source models on general benchmarks, with state-of-the-art performance on benchmarks such as CCBench, SEEDBench IMG, ScienceQA, and MMStar. The domain enhancement strategy delivers significant advantages in OCR and document understanding, validated on both public benchmarks (OCRBench 873, DocVQA 94.75%) and in-house evaluations. Notably, Qianfan-VL-8B and 70B variants incorporate long chain-of-thought capabilities, demonstrating superior performance on mathematical reasoning (MathVista 78.6%) and logical inference tasks. All models are trained entirely on Baidu's Kunlun P800 chips, validating the capability of large-scale AI infrastructure to train SOTA-level multimodal models with over 90% scaling efficiency on 5000 chips for a single task. This work establishes an effective methodology for developing domain-enhanced multimodal models suitable for diverse enterprise deployment scenarios.

cs.CV↗