Model Card for INTR: A Simple Interpretable Transformer for Fine-grained Image Classification and Analysis
INTR checkpoint on CUB dataset with backbone DETR-R50
Model Details
Model Description
Developed by: Dipanjyoti Paul, Arpita Chowdhury, Xinqi Xiong, Feng-Ju Chang, David Carlyn, Samuel Stevens, Kaiya Provost, Anuj Karpatne, Bryan Carstens, Daniel Rubenstein, Charles Stewart, Tanya Berger-Wolf, Yu Su, and Wei-Lun Chao
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
To evaluate the performance of INTR on the CUB dataset, on a multi-GPU (e.g., 4 GPUs) settings, execute the below command. INTR checkpoints are available at Fine-tune model and results.
Similarly, replace cub in the name of the checkpoint with bird or butterfly to evaluate with the Birds 525 or Cambridge Butterfly checkpoint, respectively.
If you find our work helpful for your research, please consider citing our paper as well.
@article{paul2023simple,
title={A Simple Interpretable Transformer for Fine-Grained Image Classification and Analysis},
author={Paul, Dipanjyoti and Chowdhury, Arpita and Xiong, Xinqi and Chang, Feng-Ju and Carlyn, David and Stevens, Samuel and Provost, Kaiya and Karpatne, Anuj and Carstens, Bryan and Rubenstein, Daniel and Stewart, Charles and Berger-Wolf, Tanya and Su, Yu and Chao, Wei-Lun},
journal={arXiv preprint arXiv:2311.04157},
year={2023}
}
Model Citation:
@software{Paul_A_Simple_Interpretable_2023,
author = {Paul, Dipanjyoti and Chowdhury, Arpita and Xiong, Xinqi and Chang, Feng-Ju and Carlyn, David and Stevens, Samuel and Provost, Kaiya and Karpatne, Anuj and Carstens, Bryan and Rubenstein, Daniel and Stewart, Charles and Berger-Wolf, Tanya and Su, Yu and Chao, Wei-Lun},
license = {Apache-2.0},
title = {{A Simple Interpretable Transformer for Fine-Grained Image Classification and Analysis}},
doi = {<doi once generated>},
url = {https://huggingface.co/imageomics/INTR},
version = {1.0.0},
month = sep,
year = {2023}
}
APA:
Paper:
Paul, D., Chowdhury, A., Xiong, X., Chang, F., Carlyn, D., Stevens, S., Provost, K., Karpatne, A., Carstens, B., Rubenstein, D., Stewart, C., Berger-Wolf, T., Su, Y., & Chao, W. (2023). A Simple Interpretable Transformer for Fine-Grained Image Classification and Analysis. arXiv. https://doi.org/10.48550/arXiv.2311.04157.
Model Citation:
Paul, D., Chowdhury, A., Xiong, X., Chang, F., Carlyn, D., Stevens, S., Provost, K., Karpatne, A., Carstens, B., Rubenstein, D., Stewart, C., Berger-Wolf, T., Su, Y., & Chao, W. (2023). A Simple Interpretable Transformer for Fine-Grained Image Classification and Analysis (Version 1.0.0).
Acknowledgements
Our model is inspired by the DEtection TRansformer (DETR) method.
We thank the authors of DETR for doing such great work.
The Imageomics Institute is funded by the US National Science Foundation's Harnessing the Data Revolution (HDR) program under Award #2118240 (Imageomics: A New Frontier of Biological Information Powered by Knowledge-Guided Machine Learning). Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
