A08: Identifying Birds

Using this dataset, build a convolutional neural network (CNN) that identifies a photograph of a bird as belonging to one of 200 species. (I have not completed the code myself, but I expect about 11% accuracy, based on this paper). The data is already on delenn: /bigdata/data/birds-200-caltech-ucsd-2011/. Use the photos, not the human-labeled attributes!


Your Python code (single file).

Grading rubric

  • 5 pts: All deliverables present.
  • 4 pts: All deliverables present except desired accuracy was not achieved (but it’s close).
  • 3 pts: All deliverables present except very poor accuracy (but valid CNN design).
  • 2 pts: Inappropriate dataset processing, but most code is correct.
  • 0-1 pts: Inappropriate dataset processing and most code is incorrect.

CSCI 431 material by Joshua Eckroth is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License. Source code for this website available at GitHub.