In this paper, we introduce si4onnx, a package for performing selective inference on deep learning models. Techniques such as CAM in XAI and reconstruction-based anomaly detection using VAE can be interpreted as methods for identifying significant regions within input images. However, the identified regions may not always carry meaningful significance. Therefore, evaluating the statistical significance of these regions represents a crucial challenge in establishing the reliability of AI systems. si4onnx is a Python package that enables straightforward implementation of hypothesis testing with controlled type I error rates through selective inference. It is compatible with deep learning models constructed using common frameworks such as PyTorch and TensorFlow.
@article{arxiv.2501.17415,
title = {si4onnx: A Python package for Selective Inference in Deep Learning Models},
author = {Teruyuki Katsuoka and Tomohiro Shiraishi and Daiki Miwa and Shuichi Nishino and Ichiro Takeuchi},
journal= {arXiv preprint arXiv:2501.17415},
year = {2025}
}