powered by
Infer Species for Given Detections
classify( model, detections, resize_width = 480, resize_height = 480, file_col = "filepath", crop = TRUE, normalize = TRUE, batch_size = 1, num_workers = 1, device = NULL, out_file = NULL )
detection manifest with added prediction and confidence columns
loaded classifier model
manifest of animal detections
image width input size
image height input size
column in manifest containing file paths
use bbox to crop images before feeding into model
normalize the tensor before inference
batch size for generator
number of processes
send model to the specified device
path to csv to save results to
if (FALSE) animals <- classify(classifier, animals, file_col='filepath')
Run the code above in your browser using DataLab