LIBAD

LIBAD is the dataset accompanying the paper "LIBAD: A Multimodal Anomaly Detection Benchmark for Li-Ion Battery Electrode Manufacturing".

GitHub: LIBAD

Paper: arXiv:2608.07958

The repository contains:

• LIBAD.zip: the LIBAD multimodal anomaly detection dataset.
• splits.zip: the 10 official data splits used in our experiments.

Usage

Download and extract both archives into the dataset/ directory of the official code repository:

dataset/
├── LIBAD/
└── splits/

The 10 official splits correspond to the following dataset seeds:

347, 725, 1245, 4012, 4589, 5021, 5678, 6234, 6789, 7345

Code and experimental instructions are available in the official GitHub repository:

Citation

If you use LIBAD in your research, please cite the corresponding paper:

@misc{sui2026libadmultimodalanomalydetection,
      title={LIBAD: A Multimodal Anomaly Detection Benchmark for Li-Ion Battery Electrode Manufacturing}, 
      author={Wenbo Sui and Daniel Lichau and Harold Phelippeau and Zhao Liu},
      year={2026},
      eprint={2608.07958},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2608.07958}, 
}