expected_classid = 1 # Only tigers
elif 'elp' in folder.lower() or 'elephant' in folder.lower():
expected_classid = 2
elif 'amur' in folder.lower():
expected_classid = 3
mAP = 0.
numBoxes = 0
wrong = 0
correct = 0
start = time.time()
for i in range(0, n, batchSize):
m = min(n, i + batchSize)
images = [cv2.imread(imageFiles[x]) for x in range(i, m)]
results = det.getBatchDetectionResults(images)
for j in range(i, m): # For each image in batch
imageIndex = os.path.basename(imageFiles[j]).split('.')[0]
ground_truth = box_map[imageIndex]
result = results[j - i]
if not len(result) == len(ground_truth):
print(
'Incorrect number of detections for image {}, Actual: {}, Expected: {}'
.format(imageFiles[j], len(result), len(ground_truth)))
for res in result:
if not res.classid == expected_classid:
print(
'Incorrect classification for {}, Actual class id: {}, Expected: {}'
.format(imageFiles[j], res.classid, expected_classid))
wrong += 1
else:
det.getBoundingBoxes(img)
det.getBoundingBoxes(img)
batchTimes = []
lower = 10
upper = 11
if len(args) == 4:
lower = int(args[2])
upper = int(args[3])
f = open('test_times.csv', 'w')
f.write('Batch Size,Total Time,Average Time\n')
for i in range(lower, upper, 5):
batchSize = i
# Detect batchSize images at a time
cur = 0
print('Running inference {} images at a time ...'.format(batchSize))
start = time.time()
while cur < n:
images = imageFiles[cur:min(n, cur + batchSize)]
cur += batchSize
det.getBatchDetectionResults([cv2.imread(x) for x in images])
total = time.time() - start
print('Total detection time for batchSize {}: {:.2f}s, Average: {:.2f}s'.
format(i, total, total / n))
f.write('{},{},{}\n'.format(i, total, total / n))
print('Completed...')
f.close()
