Discovering Bugs in Vision Models using Off-the-shelf Image Generation and Captioning
This work provides a proof-of-concept for using generative models to find bugs in vision models, which is incremental as it builds on existing large-scale models but offers a novel automated pipeline for failure discovery.
The authors tackled the problem of automatically discovering failures in vision models under real-world settings by leveraging off-the-shelf image-to-text and text-to-image models to generate synthetic inputs, cluster misclassifications, and identify spurious correlations, demonstrating effectiveness on ImageNet classifiers and scalability for adversarial dataset generation.
Automatically discovering failures in vision models under real-world settings remains an open challenge. This work demonstrates how off-the-shelf, large-scale, image-to-text and text-to-image models, trained on vast amounts of data, can be leveraged to automatically find such failures. In essence, a conditional text-to-image generative model is used to generate large amounts of synthetic, yet realistic, inputs given a ground-truth label. Misclassified inputs are clustered and a captioning model is used to describe each cluster. Each cluster's description is used in turn to generate more inputs and assess whether specific clusters induce more failures than expected. We use this pipeline to demonstrate that we can effectively interrogate classifiers trained on ImageNet to find specific failure cases and discover spurious correlations. We also show that we can scale the approach to generate adversarial datasets targeting specific classifier architectures. This work serves as a proof-of-concept demonstrating the utility of large-scale generative models to automatically discover bugs in vision models in an open-ended manner. We also describe a number of limitations and pitfalls related to this approach.