ContractShield: Bridging Semantic-Structural Gaps via Hierarchical Cross-Modal Fusion for Multi-Label Vulnerability Detection in Obfuscated Smart Contracts

TL;DR

ContractShield is a novel multimodal framework that enhances vulnerability detection in obfuscated smart contracts by effectively fusing semantic, temporal, and structural features through a hierarchical attention mechanism.

Contribution

It introduces a three-level fusion mechanism combining self-attention, cross-modal attention, and adaptive weighting to improve robustness against obfuscation in vulnerability detection.

Findings

Achieves 89% Hamming Score with minimal drop under obfuscation.

Detects five vulnerability types with 91% F1-score, outperforming existing methods.

Demonstrates robustness against adversarial obfuscation techniques.

Abstract

Smart contracts are increasingly targeted by adversaries employing obfuscation techniques such as bogus code injection and control flow manipulation to evade vulnerability detection. Existing multimodal methods often process semantic, temporal, and structural features in isolation and fuse them using simple strategies such as concatenation, which neglects cross-modal interactions and weakens robustness, as obfuscation of a single modality can sharply degrade detection accuracy. To address these challenges, we propose ContractShield, a robust multimodal framework with a novel fusion mechanism that effectively correlates multiple complementary features through a three-level fusion. Self-attention first identifies patterns that indicate vulnerability within each feature space. Cross-modal attention then establishes meaningful connections between complementary signals across modalities. Then, adaptive weighting dynamically calibrates feature contributions based on their reliability under obfuscation. For feature extraction, ContractShield integrates (1) CodeBERT with a sliding window mechanism to capture semantic dependencies in source code, (2) Extended long short-term memory (xLSTM) to model temporal dynamics in opcode sequences, and (3) GATv2 to identify structural invariants in control flow graphs (CFGs) that remain stable across obfuscation. Empirical evaluation demonstrates resilience of ContractShield, achieving a 89 percentage Hamming Score with only a 1-3 percentage drop compared to non-obfuscated data. The framework simultaneously detects five major vulnerability types with 91 percentage F1-score, outperforming state-of-the-art approaches by 6-15 percentage under adversarial conditions.