DeepHunter: a coverage-guided fuzz testing framework for deep neural networks

TLDR

Deep neural networks are increasingly deployed in safety‑critical domains such as autonomous driving, yet hidden defects can cause severe failures. This work introduces DeepHunter, a coverage‑guided fuzz testing framework designed to uncover defects in general‑purpose DNNs. DeepHunter employs a metamorphic mutation strategy that generates semantically preserved test inputs, uses multiple extensible coverage criteria as feedback, and combines diversity‑based and recency‑based seed selection, implementing five testing criteria and four seed strategies. Large‑scale experiments show that the metamorphic mutation achieves up to 98% validity, diversity‑based seed selection outperforms recency‑based selection, DeepHunter surpasses state‑of‑the‑art methods in coverage and defect discovery, and it effectively detects defects during DNN quantization for platform migration.

Abstract

The past decade has seen the great potential of applying deep neural network (DNN) based software to safety-critical scenarios, such as autonomous driving. Similar to traditional software, DNNs could exhibit incorrect behaviors, caused by hidden defects, leading to severe accidents and losses. In this paper, we propose DeepHunter, a coverage-guided fuzz testing framework for detecting potential defects of general-purpose DNNs. To this end, we first propose a metamorphic mutation strategy to generate new semantically preserved tests, and leverage multiple extensible coverage criteria as feedback to guide the test generation. We further propose a seed selection strategy that combines both diversity-based and recency-based seed selection. We implement and incorporate 5 existing testing criteria and 4 seed selection strategies in DeepHunter. Large-scale experiments demonstrate that (1) our metamorphic mutation strategy is useful to generate new valid tests with the same semantics as the original seed, by up to a 98% validity ratio; (2) the diversity-based seed selection generally weighs more than recency-based seed selection in boosting the coverage and in detecting defects; (3) DeepHunter outperforms the state of the arts by coverage as well as the quantity and diversity of defects identified; (4) guided by corner-region based criteria, DeepHunter is useful to capture defects during the DNN quantization for platform migration.