Selected work

Representative Papers

FastTab: A Fast Table Recognizer with a Tiny Recursive Module and 1D Transformers

May 21, 2026

This work addresses the challenge in table structure recognition of simultaneously achieving global structural consistency and precise separator localization. To this end, the authors propose an efficient grid-centric recognition approach that abandons autoregressive HTML decoding in favor of a lightweight recursive module for global reasoning. The method employs axial one-dimensional Transformers to model long-range row- and column-wise dependencies, directly predicting the number of rows and columns, header regions, and separator locations. Furthermore, it leverages ROI-aligned features to infer cross-row and cross-column relationships, enabling robust recognition of curved separators and resilience to pixel-level anonymization. Evaluated on four benchmarks—PubTabNet, FinTabNet, PubTables-1M, and SciTSR—the proposed method achieves competitive performance in both structural accuracy and inference speed.

0 citationsRead paper

TableSeq: Unified Generation of Structure, Content, and Layout

Apr 17, 2026

This work proposes the first end-to-end, single-stream autoregressive framework for table understanding that operates solely on image input, unifying structural recognition, content extraction, and cell localization within a single sequence generation process. Unlike existing approaches that decouple these tasks or rely on multi-stage pipelines and external OCR systems, the model alternately outputs HTML tags, textual content, and discrete coordinate tokens through a unified decoder. It employs a lightweight high-resolution FCN-H16 encoder, a structure prior head, and a single-layer Transformer, achieving state-of-the-art or competitive performance on benchmarks including PubTabNet (TEDS 95.23), FinTabNet, and SciTSR. The architecture significantly reduces complexity compared to multi-head or pipeline-based methods while enabling downstream applications such as index-based querying.

0 citationsRead paper

DenTab: A Dataset for Table Recognition and Visual QA on Real-World Dental Estimates

Apr 17, 2026

This study addresses the limitations of existing table recognition and visual question answering (VQA) methods, which predominantly rely on clean digital documents and struggle with noisy, structurally complex real-world administrative forms such as dental estimate sheets. To bridge this gap, the authors introduce DenTab, a dataset comprising 2,000 real-world dental estimate images annotated with high-quality HTML markup, along with 2,208 VQA tasks designed to jointly evaluate structural parsing and semantic reasoning through retrieval, aggregation, and logical consistency questions. The work further proposes Table Router Pipeline, a novel training-free framework that routes arithmetic questions to a deterministic executor, substantially enhancing inference reliability. Experiments demonstrate that current models remain weak on multi-step arithmetic and consistency tasks under realistic table layouts, whereas the proposed approach significantly improves accuracy on arithmetic VQA.

0 citationsRead paper

From Contracts to Code: Automating Smart Contract Generation with Multi-Level Finite State Machines

Jul 22, 2025

To address the high barrier to smart contract development and limited accessibility for non-technical stakeholders, this paper proposes an automated smart contract generation method based on a Multi-Level Finite State Machine (ML-FSM). The approach maps natural-language or structured traditional contracts into hierarchical, formal, and verifiable FSM models, enabling precise semantic modeling and modular decomposition. It integrates formal verification and security analysis techniques to rigorously validate functional properties and detect vulnerabilities in generated contracts. Experimental evaluation demonstrates that the framework substantially reduces development complexity while enhancing traceability, reusability, and security of smart contracts. It has successfully produced reliable, auditable contracts across diverse domains—including finance and supply chain management—thereby establishing a novel low-code paradigm for blockchain application development.

0 citationsRead paper

VISTA-OCR: Towards generative and interactive end to end OCR models

Apr 04, 2025

Traditional OCR suffers from fragmented detection and recognition pipelines, high deployment costs of vision-language large models (VLLMs), and limited support for content-driven localization and interactive tasks. To address these challenges, we propose VISTA-OCR—a lightweight, end-to-end generative OCR model. Methodologically, it introduces (1) the first prompt-controllable pretraining OCR task; (2) a unified single-branch Transformer decoder that jointly generates textual content and spatial coordinates; and (3) interactive recognition capable of handling mixed printed and handwritten text. Trained via multi-stage progressive learning, multimodal token modeling, and a synthetic–real hybrid dataset, VISTA-OCR surpasses specialized state-of-the-art methods on multiple standard benchmarks. Its 150M-parameter variant, VISTAₒₘₙᵢ, significantly reduces computational overhead while maintaining competitive accuracy and enabling efficient deployment.

0 citationsRead paper

Recent publications

Latest Papers

FastTab: A Fast Table Recognizer with a Tiny Recursive Module and 1D Transformers

May 21, 2026

This work addresses the challenge in table structure recognition of simultaneously achieving global structural consistency and precise separator localization. To this end, the authors propose an efficient grid-centric recognition approach that abandons autoregressive HTML decoding in favor of a lightweight recursive module for global reasoning. The method employs axial one-dimensional Transformers to model long-range row- and column-wise dependencies, directly predicting the number of rows and columns, header regions, and separator locations. Furthermore, it leverages ROI-aligned features to infer cross-row and cross-column relationships, enabling robust recognition of curved separators and resilience to pixel-level anonymization. Evaluated on four benchmarks—PubTabNet, FinTabNet, PubTables-1M, and SciTSR—the proposed method achieves competitive performance in both structural accuracy and inference speed.

0 citationsRead paper

TableSeq: Unified Generation of Structure, Content, and Layout

Apr 17, 2026

This work proposes the first end-to-end, single-stream autoregressive framework for table understanding that operates solely on image input, unifying structural recognition, content extraction, and cell localization within a single sequence generation process. Unlike existing approaches that decouple these tasks or rely on multi-stage pipelines and external OCR systems, the model alternately outputs HTML tags, textual content, and discrete coordinate tokens through a unified decoder. It employs a lightweight high-resolution FCN-H16 encoder, a structure prior head, and a single-layer Transformer, achieving state-of-the-art or competitive performance on benchmarks including PubTabNet (TEDS 95.23), FinTabNet, and SciTSR. The architecture significantly reduces complexity compared to multi-head or pipeline-based methods while enabling downstream applications such as index-based querying.

0 citationsRead paper

DenTab: A Dataset for Table Recognition and Visual QA on Real-World Dental Estimates

Apr 17, 2026

This study addresses the limitations of existing table recognition and visual question answering (VQA) methods, which predominantly rely on clean digital documents and struggle with noisy, structurally complex real-world administrative forms such as dental estimate sheets. To bridge this gap, the authors introduce DenTab, a dataset comprising 2,000 real-world dental estimate images annotated with high-quality HTML markup, along with 2,208 VQA tasks designed to jointly evaluate structural parsing and semantic reasoning through retrieval, aggregation, and logical consistency questions. The work further proposes Table Router Pipeline, a novel training-free framework that routes arithmetic questions to a deterministic executor, substantially enhancing inference reliability. Experiments demonstrate that current models remain weak on multi-step arithmetic and consistency tasks under realistic table layouts, whereas the proposed approach significantly improves accuracy on arithmetic VQA.

0 citationsRead paper

From Contracts to Code: Automating Smart Contract Generation with Multi-Level Finite State Machines

Jul 22, 2025

To address the high barrier to smart contract development and limited accessibility for non-technical stakeholders, this paper proposes an automated smart contract generation method based on a Multi-Level Finite State Machine (ML-FSM). The approach maps natural-language or structured traditional contracts into hierarchical, formal, and verifiable FSM models, enabling precise semantic modeling and modular decomposition. It integrates formal verification and security analysis techniques to rigorously validate functional properties and detect vulnerabilities in generated contracts. Experimental evaluation demonstrates that the framework substantially reduces development complexity while enhancing traceability, reusability, and security of smart contracts. It has successfully produced reliable, auditable contracts across diverse domains—including finance and supply chain management—thereby establishing a novel low-code paradigm for blockchain application development.

0 citationsRead paper

VISTA-OCR: Towards generative and interactive end to end OCR models

Apr 04, 2025

Traditional OCR suffers from fragmented detection and recognition pipelines, high deployment costs of vision-language large models (VLLMs), and limited support for content-driven localization and interactive tasks. To address these challenges, we propose VISTA-OCR—a lightweight, end-to-end generative OCR model. Methodologically, it introduces (1) the first prompt-controllable pretraining OCR task; (2) a unified single-branch Transformer decoder that jointly generates textual content and spatial coordinates; and (3) interactive recognition capable of handling mixed printed and handwritten text. Trained via multi-stage progressive learning, multimodal token modeling, and a synthetic–real hybrid dataset, VISTA-OCR surpasses specialized state-of-the-art methods on multiple standard benchmarks. Its 150M-parameter variant, VISTAₒₘₙᵢ, significantly reduces computational overhead while maintaining competitive accuracy and enabling efficient deployment.

0 citationsRead paper