Index-Preserving Lightweight Token Pruning for Efficient Document Understanding in Vision-Language Models

📅 2025-09-08
📈 Citations: 0
Influential: 0
📄 PDF

career value

182K/year
🤖 AI Summary
To address the high computational cost and input redundancy of vision-language models (VLMs) in document understanding, this paper proposes an index-preserving lightweight token pruning framework. Methodologically, it introduces a novel integration of a binarized block-level text classifier with a max-pooling refinement mechanism: first performing coarse-grained filtering to remove non-text background regions, then recovering fragmented text blocks via max-pooling—thereby preserving spatial coherence and maintaining original positional indices. The approach requires no fine-tuning of the backbone model and performs efficient token reduction solely during preprocessing. Experiments on real-world document datasets demonstrate significant inference acceleration (≈40% FLOPs reduction) while retaining baseline accuracy on key tasks such as OCR-enhanced visual question answering. The core contribution is the first realization of high-fidelity, index-aligned token pruning for document images—establishing a practical new paradigm for deploying VLMs under resource constraints.

Technology Category

Application Category

📝 Abstract
Recent progress in vision-language models (VLMs) has led to impressive results in document understanding tasks, but their high computational demands remain a challenge. To mitigate the compute burdens, we propose a lightweight token pruning framework that filters out non-informative background regions from document images prior to VLM processing. A binary patch-level classifier removes non-text areas, and a max-pooling refinement step recovers fragmented text regions to enhance spatial coherence. Experiments on real-world document datasets demonstrate that our approach substantially lowers computational costs, while maintaining comparable accuracy.
Problem

Research questions and friction points this paper is trying to address.

Reduces computational costs in document understanding VLMs
Filters non-informative background regions from documents
Maintains accuracy while pruning lightweight tokens
Innovation

Methods, ideas, or system contributions that make the work stand out.

Lightweight token pruning framework filters background
Binary patch classifier removes non-text areas
Max-pooling refinement recovers fragmented text regions