Four CSV files for each project/PID. CSVs will be copied to Wildlife Insights’ GCP to the specified Bucket using the Base Path: wildlife_insights_bulk_uploads/<PID>/ where <PID> is replaced by the plot, subproject, or project being uploaded. 

Records will be flattened in the image.csv file. This means that if a single image contains more than one species there will be one row in the CSV for each species (e.g., If an image shows two raccoons and one deer, the CSV will contain two rows: the first showing two raccoons and the second showing the single deer). 

see for samples:

https://docs.google.com/spreadsheets/d/1iEcHs0Y49W5hx7aoMSFge_1-Q_VfMdl8d56x27heuNY/edit#gid=134657300

https://docs.google.com/spreadsheets/d/1l6gVVOHnky-e5vTPP7wCmX5-0_rrdOv51hDEEFOPyng/edit#gid=134657300

https://github.com/ConservationInternational/Wildlife-Insights----Data-Migration/tree/master/Datasets/Example_Dataset/CafeFauna/Cafe_Fauna_wi_batch_upload