Have a question about dlib usage.

[JainJang]

I'm trying to train my own aging data set and want to crop the frontal faces almost 30,000 images.
I input the command as below:
$python guess.py --model_type inception --model_dir model --filename dlib1.csv --face_detection_type dlib --face_detection_model shape_predictor_68_face_landmarks.dat

In the dlib1.csv, I have image data path.

And I have an error like this:

Using face detector (dlib) shape_predictor_68_face_landmarks.dat
dlib1.csv
OpenCV Error: Assertion failed (scn == 3 || scn == 4) in ipp_cvtColor, file /builds/bob/bob.conda/conda-env/opencv_1489404393345/work/opencv-3.1.0/modules/imgproc/src/color.cpp, line 7456
Traceback (most recent call last):
File "guess.py", line 207, in
tf.app.run()
File "/home/anaconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow/python/platform/app.py", line 124, in run
_sys.exit(main(argv))
File "guess.py", line 140, in main
face_files, rectangles = face_detect.run(FLAGS.filename)
File "/home/python/age/dlibdetect.py", line 16, in run
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
cv2.error: /builds/bob/bob.conda/conda-env/opencv_1489404393345/work/opencv-3.1.0/modules/imgproc/src/color.cpp:7456: error: (-215) scn == 3 || scn == 4 in function ipp_cvtColor

How can I use the guess.py for face images cropping with big dataset?