ValueError: operands could not be broadcast together with shapes (2,1,128) (413,369,3)

• face_recognition version:1.2.3
• Python version: 3.5.4
• Operating System: Windows 10 (x64)

Description

Hi everyone. I am building a facial recognition software that is supposed to identify faces with openCV, take the identified faces, save them as a .jpg file and use that said file for comparison, however there seemed to be some complication when I use the face_recognition.compare_faces function.

What I Did

` def getMatches(self): matches = [] encodings = [] name = “” faceNames = [ “Jass”,“Neo” ] faceToCheck = fr.load_image_file(“facelock.jpg”) faceEncoding = fr.face_encodings(faceToCheck)[0]

    face2 = fr.load_image_file("Neo.jpg") 
    encodings.append(fr.face_encodings(face2)[0])
    
    matches = fr.compare_faces(encodings, faceEncoding)
    name = "UNKNOWN"

    if True in matches:
        ind = matches.index(True)
        name = faceNames[ind]
        print("The name is:" +str(name))

` I am getting this: Traceback (most recent call last): File “C:\Users\Janda\Desktop\MUKA FINAL\MukaMain.py”, line 201, in getFace self.getMatches() File “C:\Users\Janda\Desktop\MUKA FINAL\MukaMain.py”, line 219, in getMatches matches = fr.compare_faces(encodings, faceEncoding) File “C:\Users\Janda\AppData\Local\Programs\Python\Python35-32\lib\site-packages\face_recognition-1.2.3-py3.5.egg\face_recognition\api.py”, line 222, in compare_faces return list(face_distance(known_face_encodings, face_encoding_to_check) <= tolerance) File “C:\Users\Janda\AppData\Local\Programs\Python\Python35-32\lib\site-packages\face_recognition-1.2.3-py3.5.egg\face_recognition\api.py”, line 72, in face_distance return np.linalg.norm(face_encodings - face_to_compare, axis=1)

ValueError: operands could not be broadcast together with shapes (1,128) (103,103,3)

Here are the images: Facelock Neo

Thank you for your help 😃

[Gangeya003]

The issue is when we encode images with improper light on your face, the method face_encoding returns an empty list and when you compare this with that new image’s encoding this error occurs… this might even occur in the reverse case.

[root-Akshay]

@NeoDaenium @ageitgey I am getting this error , tried to resolve by looking some solution mentioned on web but could’nt make it work :

import face_recognition import os import cv2

KNOWN_FACES_DIR = “known_faces” UNKNOWN_FACES_DIR = “unknown_faces” TOLERANCE = 0.6 FRAME_THICKNESS = 3 FONT_THICKNESS = 2 MODEL = “cnn” #hog for cpu, cnn works slow on cpu

print(“Loading known faces”)

known_faces = [] known_names = []

for name in os.listdir(KNOWN_FACES_DIR): for filename in os.listdir(f"{KNOWN_FACES_DIR}/{name}“): image = face_recognition.load_image_file(f”{KNOWN_FACES_DIR}/{name}/{filename}") # encoding = face_recognition.face_encodings(image)[0] encoding = face_recognition.face_encodings(image) known_faces.append(encoding) # print(known_faces) known_names.append(name) print(known_names)

print("processing unknown faces ")

for filename in os.listdir(UNKNOWN_FACES_DIR): print(filename) image = face_recognition.load_image_file(f"{UNKNOWN_FACES_DIR}/{filename}") locaations = face_recognition.face_locations(image, model=MODEL) encodings = face_recognition.face_encodings(image, locaations) image = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)

for face_encoding, face_location in zip(encodings, locaations):
    results = face_recognition.compare_faces(known_faces, face_encoding, TOLERANCE)
    print(results)
    match = None
    if True in results:
        match = known_names[results.index(True)]
        print(f"Match Found : {match}")

        top_left = (face_location[3], face_location[0])
        bottom_right = (face_location[1], face_location[2])

        color = [0, 255, 0]

        cv2.rectangle(image, top_left, bottom_right, color, FRAME_THICKNESS)

        top_left = (face_location[3], face_location[2])
        bottom_right = (face_location[1], face_location[2]+22)
        cv2.rectangle(image, top_left, bottom_right, color, cv2.FILLED)
        cv2.putText(image, match, (face_location[2]+15), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (200,200,200), FONT_THICKNESS)


cv2.imshow(filename, image)
cv2.waitKey(0)
cv2.destroyWindow(filename)

The problem lies over here, encoding = face_recognition.face_encodings(image) known_faces.append(encoding)

Instead of appending encoded list into another list, directly pass the encoding to face_recognition.compare_faces()

known_encoding = face_recognition.face_encodings(image) results = face_recognition.compare_faces(known_encoding, face_encoding, TOLERANCE)

I get what you are trying to achieve with this code, for a rectangle border to be shown on the image:

cv2.rectangle(image, top_left, bottom_right, color, cv2.BORDER_WRAP)