Skip to content

code and link to the dataset for Kenyan Food detection paper accepted as a paper in MADiMA 2019 Workshop as part of ACM MM 2019 conference.

Notifications You must be signed in to change notification settings

monajalal/Kenyan-Food

Repository files navigation

Scraping Social Media Photos Posted in Kenya and Elsewhere to Detect and Analyze Food Types

M. Jalal*, K. Wang*, J. Sankara, Y. Zheng, E. O. Nsoesie, and M. Betke. Scraping Social Media Photos Posted in Kenya and Elsewhere to Detect and Analyze Food Types. published in MADiMA 2019, the 5th International Workshop on Multimedia Assisted Dietary Management. In conjunction with the 27th ACM International Conference on Multimedia (ACMMMM 2019) Nice, France, October 21, 2019. 10 pages.

Kenyan food/non food dataset

Kenya104K Image Samples

Kenyan104K dataset: https://www.dropbox.com/scl/fi/ie3ddguh06t0uxkqlmfln/KFD.zip?rlkey=vt4xzclx5y1im7433b9vqlv9r&dl=0

Kenyan Food Type Dataset (KenyanFood13)

KenyanFood13 Image Samples

KenyanFood13 dataset: https://www.dropbox.com/scl/fi/hk1llnnv6bpjw153epfxo/Food13.zip?rlkey=o7iq83g4g0xjeif45ibxd9kkb&dl=0

Link to ArXiv paper: https://arxiv.org/abs/1909.00134

Funding

This research was partially funded by the following awards:

  • NSF Award #1838193 BIGDATA: IA: Multiplatform, Multilingual, and Multimodal Tools for Analyzing Public Communication in over 100 Languages

  • Hariri Institute for Computing and Computational Science & Engineering at Boston University

Credits/Authors

Kaihong Wang*, Mona Jalal*, Jefferson Sankara, Yi Zheng, Elaine Nsoesie, Margrit Betke

BibTeX for Citing

@inproceedings{JalalWaJeZhNsBe19,
  title={Scraping social media photos posted in Kenya and elsewhere to detect and analyze food types},
  author={Jalal, Mona and Wang, Kaihong and Jefferson, Sankara and Zheng, Yi and Nsoesie, Elaine O and Betke, Margrit},
  booktitle={Proceedings of the 5th International Workshop on Multimedia Assisted Dietary Management},
  pages={50--59},
  year={2019}
}

Acknowledgement

We are grateful to Shuai Wei for extensive support on getting to run our code work on SCC as well as helping us with various troubleshooting for both our PHP-based scraping and DNN codes. Shuai can be reached at [email protected]

TODO

  • Adding the 5 fold cross-validation folders for Kenyan food/non-food classifier as well as Kenyan Food Type Recognizer in Google Drive

  • Adding documentation how to run the scraping script with a minimal example for both hashtag-based as well as location-based scenarios.

  • Adding documentation for running the Kenyan Food/Non-Food Classifier as well as Kenyan Food Type Recognizer both in Shared Cluster Center (SCC@BU) as well as on a Deep Learning station.

  • Adding a requirement.txt that shows all the necessary packages for this project.