Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.22413 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914587904835584 |
|---|---|
| author | Wang, Yandi Zhan, Libin Huang, Ziwei Luo, Tiancheng Jiang, Yuxuan Dong, Wang Gan, Leilei Chen, Jun |
| author_facet | Wang, Yandi Zhan, Libin Huang, Ziwei Luo, Tiancheng Jiang, Yuxuan Dong, Wang Gan, Leilei Chen, Jun |
| contents | Extracting structured information from visual documents (Visual Information Extraction, VIE) is a cornerstone of business automation. While recent Multimodal Large Language Models (MLLMs) have shown promising capabilities, existing benchmarks suffer from critical limitations in scale and realism, lack semantic granularity, and fail to cover diverse document types. To bridge this gap, we introduce ReceiptBench, a large-scale, human-annotated benchmark consisting of 10k diverse receipts, organizing information extraction into four hierarchical sub-tasks: (1) Basic Perception for raw text spotting, (2) Format Normalization for strictly following standardization instructions, (3) Semantic Reasoning for inferring implicit attributes from context, and (4) Structure Parsing for handling nested line items. Furthermore, we propose a two-stage training framework incorporating Metric-Aware Group Relative Policy Optimization (GRPO), which translates rigorous evaluation constraints into reinforcement learning signals to enhance structural consistency. Extensive experiments demonstrate that our method yields state-of-the-art performance, surpassing leading proprietary models on complex reasoning tasks. We release our datasets and code at https://github.com/wwwT0ri/ReceiptBench. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_22413 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | From Recognition to Reasoning: Benchmarking and Enhancing MLLMs on Real-World Receipt Document Understanding Wang, Yandi Zhan, Libin Huang, Ziwei Luo, Tiancheng Jiang, Yuxuan Dong, Wang Gan, Leilei Chen, Jun Computer Vision and Pattern Recognition Extracting structured information from visual documents (Visual Information Extraction, VIE) is a cornerstone of business automation. While recent Multimodal Large Language Models (MLLMs) have shown promising capabilities, existing benchmarks suffer from critical limitations in scale and realism, lack semantic granularity, and fail to cover diverse document types. To bridge this gap, we introduce ReceiptBench, a large-scale, human-annotated benchmark consisting of 10k diverse receipts, organizing information extraction into four hierarchical sub-tasks: (1) Basic Perception for raw text spotting, (2) Format Normalization for strictly following standardization instructions, (3) Semantic Reasoning for inferring implicit attributes from context, and (4) Structure Parsing for handling nested line items. Furthermore, we propose a two-stage training framework incorporating Metric-Aware Group Relative Policy Optimization (GRPO), which translates rigorous evaluation constraints into reinforcement learning signals to enhance structural consistency. Extensive experiments demonstrate that our method yields state-of-the-art performance, surpassing leading proprietary models on complex reasoning tasks. We release our datasets and code at https://github.com/wwwT0ri/ReceiptBench. |
| title | From Recognition to Reasoning: Benchmarking and Enhancing MLLMs on Real-World Receipt Document Understanding |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.22413 |