GIZ NLP Agricultural Keyword Spotter

The objective of this competition is to build a machine learning model to identify the agricultural keyword (which may be in English or Luganda) spoken in an audio clip. The keywords relate to crops, diseases, fertilizers, herbicides or other general agricultural topics.

This solution will help researchers from Makerere University who are developing a speech recognition model to automate the process of monitoring Luganda radio programs for agriculture-related information. This solution will enable more efficient monitoring and analysis of local language radio programs, and your work can possibly open doors for this type of natural language processing tasks in other local languages across Africa and across other sectors that use radio as a means of communication.


Competition closes on 29 November 2020.

Final submissions must be received by 11:59 PM GMT.


  • 1st Place: $5,000 USD

  • 2nd Place: $1,500 USD

  • 3rd Place: $500 USD

Join Competition Here