MosAIc: Finding Artistic Connections across Culture with Conditional Image Retrieval

Mark Hamilton, Stephanie Fu, Mindren Lu, Johnny Bui, Darius Bopp, Zhenbang Chen, Felix Tran, Margaret Wang, Marina Rogers, Lei Zhang, Chris Hoder, William T. Freeman
Proceedings of the NeurIPS 2020 Competition and Demonstration Track, PMLR 133:133-155, 2021.

Abstract

We introduce MosAIc, an interactive web app that allows users to find pairs of semantically related artworks that span different cultures, media, and millennia. To create this application, we introduce Conditional Image Retrieval (CIR) which combines visual similarity search with user supplied filters or “conditions”. This technique allows one to find pairs of similar images that span distinct subsets of the image corpus. We provide a generic way to adapt existing image retrieval data-structures to this new domain and provide theoretical bounds on our approach’s efficiency. To quantify the performance of CIR systems, we introduce new datasets for evaluating CIR methods and show that CIR performs non-parametric style transfer. Finally, we demonstrate that our CIR data-structures can identify “blind spots” in Generative Adversarial Networks (GAN) where they fail to properly model the true data distribution.

Cite this Paper

BibTeX
@InProceedings{pmlr-v133-hamilton21a, title = {MosAIc: Finding Artistic Connections across Culture with Conditional Image Retrieval}, author = {Hamilton, Mark and Fu, Stephanie and Lu, Mindren and Bui, Johnny and Bopp, Darius and Chen, Zhenbang and Tran, Felix and Wang, Margaret and Rogers, Marina and Zhang, Lei and Hoder, Chris and Freeman, William T.}, booktitle = {Proceedings of the NeurIPS 2020 Competition and Demonstration Track}, pages = {133--155}, year = {2021}, editor = {Escalante, Hugo Jair and Hofmann, Katja}, volume = {133}, series = {Proceedings of Machine Learning Research}, month = {06--12 Dec}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v133/hamilton21a/hamilton21a.pdf}, url = {https://proceedings.mlr.press/v133/hamilton21a.html}, abstract = {We introduce MosAIc, an interactive web app that allows users to find pairs of semantically related artworks that span different cultures, media, and millennia. To create this application, we introduce Conditional Image Retrieval (CIR) which combines visual similarity search with user supplied filters or “conditions”. This technique allows one to find pairs of similar images that span distinct subsets of the image corpus. We provide a generic way to adapt existing image retrieval data-structures to this new domain and provide theoretical bounds on our approach’s efficiency. To quantify the performance of CIR systems, we introduce new datasets for evaluating CIR methods and show that CIR performs non-parametric style transfer. Finally, we demonstrate that our CIR data-structures can identify “blind spots” in Generative Adversarial Networks (GAN) where they fail to properly model the true data distribution.} }
Endnote
%0 Conference Paper %T MosAIc: Finding Artistic Connections across Culture with Conditional Image Retrieval %A Mark Hamilton %A Stephanie Fu %A Mindren Lu %A Johnny Bui %A Darius Bopp %A Zhenbang Chen %A Felix Tran %A Margaret Wang %A Marina Rogers %A Lei Zhang %A Chris Hoder %A William T. Freeman %B Proceedings of the NeurIPS 2020 Competition and Demonstration Track %C Proceedings of Machine Learning Research %D 2021 %E Hugo Jair Escalante %E Katja Hofmann %F pmlr-v133-hamilton21a %I PMLR %P 133--155 %U https://proceedings.mlr.press/v133/hamilton21a.html %V 133 %X We introduce MosAIc, an interactive web app that allows users to find pairs of semantically related artworks that span different cultures, media, and millennia. To create this application, we introduce Conditional Image Retrieval (CIR) which combines visual similarity search with user supplied filters or “conditions”. This technique allows one to find pairs of similar images that span distinct subsets of the image corpus. We provide a generic way to adapt existing image retrieval data-structures to this new domain and provide theoretical bounds on our approach’s efficiency. To quantify the performance of CIR systems, we introduce new datasets for evaluating CIR methods and show that CIR performs non-parametric style transfer. Finally, we demonstrate that our CIR data-structures can identify “blind spots” in Generative Adversarial Networks (GAN) where they fail to properly model the true data distribution.
APA
Hamilton, M., Fu, S., Lu, M., Bui, J., Bopp, D., Chen, Z., Tran, F., Wang, M., Rogers, M., Zhang, L., Hoder, C. & Freeman, W.T.. (2021). MosAIc: Finding Artistic Connections across Culture with Conditional Image Retrieval. Proceedings of the NeurIPS 2020 Competition and Demonstration Track, in Proceedings of Machine Learning Research 133:133-155 Available from https://proceedings.mlr.press/v133/hamilton21a.html.

Related Material