wrong datatype for eyesCenter in facealiner #254

Activity

1. andvikt

   added a commit that references this issue on Jul 27, 2021

   fix PyImageSearch#254

   This issue will close once commit 6928183 is merged into the 'master' branch.

   6928183
2. Wingscape commented on Dec 28, 2021

   

   Wingscape

   on Dec 28, 2021

   More actions

   Wow that's something why i can't fix this bug whenever i search on google. I think im gonna use haar cascade if this issue doesn't solve yet
3. ariG23498 commented on Jan 27, 2022

   

   ariG23498

   on Jan 27, 2022

   More actions

   Hey folks,

   Thanks for the issue. Could you give us a reproducible bug to check? A colab notebook with the bug would do! TIA 😄
4. akapsch commented on Apr 13, 2022

   

   akapsch

   on Apr 13, 2022

   More actions

   I can confirm that this is still an issue. I was following the face align tutorial, and kept running into this exact issue. I eventually added all of the facealigner.py code into my code and got the same error. Adding int() to the tuple fixed it:

   eyesCenter = (int((leftEyeCenter[0] + rightEyeCenter[0]) // 2),
           int((leftEyeCenter[1] + rightEyeCenter[1]) // 2))
   

   Maybe I'm using a bad version? This is what my pip reports:

   Package               Version
   --------------------- --------
   cmake                 3.22.3
   dlib                  19.23.1
   imutils               0.5.4
   numpy                 1.22.3
   opencv-contrib-python 4.5.5.64
   

   👍React with 👍9Reacted by Enikő Villantová, Ollie Graham, Eduardo Sales Castelo Branco, Aritra Roy Gosthipaty, Sangeen Khan, mahmoud alforawi, Abhishek Thanki, TigerKim9 and Azure-sco
5. vikaskookna

   assigned

   ariG23498

   on Jul 14, 2022
6. vikaskookna commented on Jul 14, 2022

   

   vikaskookna

   on Jul 14, 2022

   More actions

   Hi @ariG23498 this is still an issue as I got a ticket from a customer today about this
7. ariG23498

   assigned

   abhishekthanki

   and

   ritwikraha

   on Jul 19, 2022
8. ariG23498 commented on Jul 19, 2022

   

   ariG23498

   on Jul 19, 2022

   More actions

   Tagging @abhishekthanki and @ritwikraha to this issue!
9. ariskoutris commented on Nov 2, 2022

   

   ariskoutris

   on Nov 2, 2022

   Last edited by ariskoutris

   More actions

   I had this issue too with opencv-python 4.6.0.66 and numpy 1.19.5. Thanks @AKSoapy29 for the easy fix.
10. BilalAlsharif commented on Nov 6, 2022

    

    BilalAlsharif

    on Nov 6, 2022

    More actions

    det = dlib.get_frontal_face_detector()
    pred = dlib.shape_predictor(r'...\shape_predictor_68_face_landmarks.dat')
    alg = FaceAligner(pred, desiredFaceWidth = 256)
    This tries to do face-normalisation

    This makes this error 'Can't parse 'Center''

    frame = cv2.imread('img.jpg')
    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    rect = detector(gray,1)[0]
    faceAligned = aligner.align(frame, gray, rect)

    File ~\AppData\Local\Programs\Python\Python310\lib\site-packages\imutils\face_utils\facealigner.py:68, in FaceAligner.align(self, image, gray, rect)
    64 eyesCenter = ((leftEyeCenter[0] + rightEyeCenter[0]) // 2,
    65 (leftEyeCenter[1] + rightEyeCenter[1]) // 2)
    67 # grab the rotation matrix for rotating and scaling the face
    ---> 68 M = cv2.getRotationMatrix2D(eyesCenter, angle, scale)
    70 # update the translation component of the matrix
    71 tX = self.desiredFaceWidth * 0.5

    TypeError: Can't parse 'center'. Sequence item with index 0 has a wrong type
11. dunland commented on Dec 25, 2022

    

    dunland

    on Dec 25, 2022

    More actions

    For me, casting the eyesCenter from <class 'numpy.int64> to type <class 'int'> instead of fixed it:

    in faceAligner.py:64 I did:

    eyesCenter = (int((leftEyeCenter[0] + rightEyeCenter[0]) // 2),
    		int((leftEyeCenter[1] + rightEyeCenter[1]) // 2))
12. erfan-zekri commented on May 22, 2024

    

    erfan-zekri

    on May 22, 2024

    More actions

    For me, casting the eyesCenter from <class 'numpy.int64> to type <class 'int'> instead of fixed it:

    in faceAligner.py:64 I did:

    eyesCenter = (int((leftEyeCenter[0] + rightEyeCenter[0]) // 2),
    		int((leftEyeCenter[1] + rightEyeCenter[1]) // 2))

    This works for me! Thanks @dunland
13. abhishekthanki commented on May 22, 2024

    

    abhishekthanki

    on May 22, 2024 via email

    More actions

    Hi Pyimagesearch/imutils, We know how annoying auto-reply and email confirmation responses are, but we wanted to send you a quick reply and let you know that your email has safely landed in our inbox. Hooray! We try to respond to emails within 72 hours, Monday-Friday. In the meantime, here’s a few quick tips to help you out: 1. *We prioritize customer emails.* Full transparency, we receive tens of thousands of emails a year, and we prioritize customer support emails. If you’d like to become a customer, you can see our full catalog of books and courses here <https://www.pyimagesearch.com/books-and-courses/> . 2. *Not sure which of our books/courses is right for you? Need a personalized recommendation from us?* No problem, we can get you unblocked and moving forward in < 13.84 seconds (*seriously, we’ve tested it!*) To start, take this short multiple choice survey <https://www.pyimagesearch.com/survey/> — it will compile your answers and recommend our book/course best suited for you and your needs. 3. *If you are not a customer and not interested in becoming a customer, you probably won’t get a response.* We do our best to respond to all emails we receive, but there’s only so much time in the day to offer free help and support and we continue to get more and more emails. Therefore, we have elected to “do the most good” with our limited resources and serve the computer vision and deep learning community at large by publishing weekly, high-quality (and did we mention *free*) tutorials. We hope these give you everything you need to be successful with your project or research. 4. *If you are a customer, we can only help with questions directly related to my books, courses, and blog posts.* If you have a question related to a topic covered in our books/courses, or have an issue with code error’ing out, we’ll be more than happy to help. Otherwise, if you need targeted 1:1 help with your specific project, we've partnered with Dr. Tim Oates, who offers office hours and consulting on behalf of PyImageSearch. Here’s a quick survey <https://www.pyimagesearch.com/consulting/> you can fill out to start the office hours/consulting request with Tim. We really appreciate your understanding, we hope this email didn’t let you down. See you at pyimagesearch.com! *PyImageSearch Team* {#HS:2604959745-309633#}
14. SanjithKumar2 commented on Nov 20, 2025

    

    SanjithKumar2

    on Nov 20, 2025

    More actions

    Why has the issue not been corrected still it's nearly 4 years...
15. abhishekthanki commented on Nov 20, 2025

    

    abhishekthanki

    on Nov 20, 2025 via email

    More actions

    Hi Pyimagesearch/imutils, We know how annoying auto-reply and email confirmation responses are, but we wanted to send you a quick reply and let you know that your email has safely landed in our inbox. Hooray! We try to respond to emails within 72 hours, Monday-Friday. In the meantime, here’s a few quick tips to help you out: 1. *We prioritize customer emails.* Full transparency, we receive tens of thousands of emails a year, and we prioritize customer support emails. If you’d like to become a customer, you can see our full catalog of books and courses here <https://www.pyimagesearch.com/books-and-courses/> . 2. *Not sure which of our books/courses is right for you? Need a personalized recommendation from us?* No problem, we can get you unblocked and moving forward in < 13.84 seconds (*seriously, we’ve tested it!*) To start, take this short multiple choice survey <https://www.pyimagesearch.com/survey/> — it will compile your answers and recommend our book/course best suited for you and your needs. 3. *If you are not a customer and not interested in becoming a customer, you probably won’t get a response.* We do our best to respond to all emails we receive, but there’s only so much time in the day to offer free help and support and we continue to get more and more emails. Therefore, we have elected to “do the most good” with our limited resources and serve the computer vision and deep learning community at large by publishing weekly, high-quality (and did we mention *free*) tutorials. We hope these give you everything you need to be successful with your project or research. 4. *If you are a customer, we can only help with questions directly related to my books, courses, and blog posts.* If you have a question related to a topic covered in our books/courses, or have an issue with code error’ing out, we’ll be more than happy to help. Otherwise, if you need targeted 1:1 help with your specific project, we've partnered with Dr. Tim Oates, who offers office hours and consulting on behalf of PyImageSearch. Here’s a quick survey <https://www.pyimagesearch.com/consulting/> you can fill out to start the office hours/consulting request with Tim. We really appreciate your understanding, we hope this email didn’t let you down. See you at pyimagesearch.com! *PyImageSearch Team* {#HS:3145351818-332595#}