42 lines
1.1 KiB
Python
42 lines
1.1 KiB
Python
from NeuralNetwork import prediction
|
|
import os
|
|
|
|
def test_all():
|
|
material = 'glass'
|
|
dir = "./resources/trash_dataset/test/" + material
|
|
for i in range(0, 56, 5):
|
|
count = 0
|
|
full_count = 0
|
|
for file in os.listdir(dir):
|
|
full_count += 1
|
|
path = os.path.join(dir, file)
|
|
# print(path)
|
|
result = prediction.getPrediction(path, 'trained_nn_'+str(i)+'.pth')
|
|
if result == material:
|
|
count += 1
|
|
print('siec ' + str(i) + ': ' + str(count) + '/' + str(full_count))
|
|
|
|
def test_one():
|
|
network = 20
|
|
material = 'paper'
|
|
dir = "./resources/trash_dataset/test/" + material
|
|
count = 0
|
|
full_count = 0
|
|
for file in os.listdir(dir):
|
|
full_count += 1
|
|
path = os.path.join(dir, file)
|
|
result = prediction.getPrediction(path, 'trained_nn_'+str(network)+'.pth')
|
|
if result == material:
|
|
count += 1
|
|
else:
|
|
print(path)
|
|
print('siec ' + str(network) + ': ' + str(count) + '/' + str(full_count))
|
|
|
|
|
|
def main():
|
|
test_one()
|
|
|
|
|
|
|
|
if __name__ == '__main__':
|
|
main() |