Face_recognition: IndexError: list index out of range

[wanesta]

IndexError: list index out of range

my code:

import face_recognition
known_image = face_recognition.load_image_file("D:/1.jpg")
unknown_image = face_recognition.load_image_file("D:/2.jpg")
biden_encoding = face_recognition.face_encodings(known_image)[0]

[wanesta]

I'm run code with windows

[ageitgey]

If no face is found in the image, the encodings array will be empty. So check the length of the array before you try to access the first element (i.e. element [0]) of the array.

For example:

`python encodings = face_recognition.face_encodings(known_image) if len(encodings) > 0: biden_encoding = encodings[0] else: print("No faces found in the image!") quit()

[wanesta]

My image is my dog,I want to recognition dog face. How to do?

[wanesta]

thak you very much.

[spad]

Hi ageitgey,
sometimes it happens that face_locations finds faces, but face_encodings not!
How it's possibile?
I attached the cropped face from rect returned by face_locations. The same image then fails in face_encodings.
I'm using a client/server structure:
the client looks for faces in camera frames, crops images and send them to the server.
Then the server accomplishes face recognition encodings face pictures.

So, the face_encodings functions receives just the cropped face picture and not the entire frame.

I tried both hog and cnn model in face_locations with the same result.

[ageitgey]

@spad if you are only sending the cropped area of the face, you need to tell it that the whole image contains the face. If you don't, it will search for the face within the image and won't know the whole image itself is the face. If it can't find a face inside the image, it won't be able to do anything.

So just tell it that the whole image is the face:

import face_recognition

img = face_recognition.load_image_file("your_cropped_image")

# Assume the whole image is the location of the face
height, width, _ = img.shape
# location is in css order - top, right, bottom, left
face_location = (0, width, height, 0)

encodings = face_recognition.face_encodings(img, known_face_locations=[face_location])

Note: I didn't really test this code, so just double check that the results it gives you are correct.

[spad]

@ageitgey
I solved now just oversizing the cropped face before face_encodings.
But I'm interested to you solutions, I'will try. Thanks.

[ageitgey]

@SaraHatem check this out - https://github.com/ageitgey/face_recognition/wiki/Common-Errors#list-index-out-of-range-errors

BTW this library won't detect cat faces, only human faces.

[SaraHatem]

@ageitgey ok thanks alot, but do u have any recommendations about an animal or cat and dog facial recognition library or software i can use in my project, i've been searching alot but nothing works.
thanks alot inadvance

[ageitgey]

Unfortunately I don't, sorry!

[HAKANMAZI]

@spad if you are only sending the cropped area of the face, you need to tell it that the whole image contains the face. If you don't, it will search for the face within the image and won't know the whole image itself is the face. If it can't find a face inside the image, it won't be able to do anything.

So just tell it that the whole image is the face:

import face_recognition

img = face_recognition.load_image_file("your_cropped_image")

# Assume the whole image is the location of the face
height, width, _ = img.shape
# location is in css order - top, right, bottom, left
face_location = (0, width, height, 0)

encodings = face_recognition.face_encodings(img, known_face_locations=[face_location])

Note: I didn't really test this code, so just double check that the results it gives you are correct.

I tested it works, Thanks

[Rajatkul1998]

just remove the 0 in biden_encoding = face_recognition.face_encodings(known_image)[0]
if you are passing more than 1 image

[panchami88]

BUT IF WE RMOVE 0, ITS GIVING ERROR LIKE THIS
ValueError: operands could not be broadcast together with shapes (1,2) (128,)

[amanmishra-edunomics]

Sir, I m get error list Index out of range, whenever someone is capturing the image and after passing it to that api.
def facedect(loc): #Here loc is location of my captured image
cam = cv2.VideoCapture(0)
s, img = cam.read()
if s:
face_1_image = face_recognition.load_image_file(loc)
face_1_face_encoding = face_recognition.face_encodings(face_1_image)[0]

    # face_1_face_encoding=face_1_face_encoding[0]

    #

    small_frame = cv2.resize(img, (0, 0), fx=0.25, fy=0.25)

    rgb_small_frame = small_frame[:, :, ::-1]

    face_locations = face_recognition.face_locations(rgb_small_frame)
    face_encodings = face_recognition.face_encodings(rgb_small_frame, face_locations)
    if len(face_encodings)>0:
        face_encodings = face_encodings[0]
        check = face_recognition.compare_faces([face_1_face_encoding], face_encodings)
        print(check)
        if check[0]:
            return True
        else:
            return False
    else:
        print("No face Found")
        quit()

and the code i m using for camera

def camera1():
# Camera 0 is the integrated web cam on my netbook
camera_port = 0

# Number of frames to throw away while the camera adjusts to light levels
ramp_frames = 30
# global Name
# Fetching the name of the person
Dir = 'E:\Practice\Django\Facebook2\Camera\image.jpg'

# Now we can initialize the camera capture object with the cv2.VideoCapture class.
# All it needs is the index to a camera port.
camera = cv2.VideoCapture(camera_port)

# Captures a single image from the camera and returns it in PIL format
def get_image():
    # read is the easiest way to get a full image out of a VideoCapture object.
    retval, im = camera.read()
    return im

# Ramp the camera - these frames will be discarded and are only used to allow v4l2
# to adjust light levels, if necessary
for i in range(ramp_frames):
    temp = get_image()
print("Taking image...")
# Take the actual image we want to keep
camera_capture = get_image()
file = Dir
# A nice feature of the imwrite method is that it will automatically choose the
# correct format based on the file extension you provide. Convenient!
cv2.imwrite(file, camera_capture)
# cv2.imwrite(UserProfiles.objects.create(extra_ident=id),camera_capture)

# You'll want to release the camera, otherwise you won't be able to create a new
# capture object until your script exits
del camera
return file

[Tareqtaleb45]

IndexError: liste des index hors limites
mon code est :
import cv2
import dlib
import PIL.Image
import numpy as np
from imutils import face_utils
import argparse
from pathlib import Path
import os
import ntpath

print('[INFO] Starting System...')
print('[INFO] Importing pretrained model..')
pose_predictor_68_point = dlib.shape_predictor("pretrained_model/shape_predictor_68_face_landmarks.dat")
pose_predictor_5_point = dlib.shape_predictor("pretrained_model/shape_predictor_5_face_landmarks.dat")
face_encoder = dlib.face_recognition_model_v1("pretrained_model/dlib_face_recognition_resnet_model_v1.dat")
face_detector = dlib.get_frontal_face_detector()
print('[INFO] Importing pretrained model..')
def transform(image, face_locations):
coord_faces = []
for face in face_locations:
rect = face.top(), face.right(), face.bottom(), face.left()
coord_face = max(rect[0], 0), min(rect[1], image.shape[1]), min(rect[2], image.shape[0]), max(rect[3], 0)
coord_faces.append(coord_face)
return coord_faces
def encode_face(image):
face_locations = face_detector(image, 1)
face_encodings_list = []
landmarks_list = []
for face_location in face_locations:
# DETECT FACES
shape = pose_predictor_68_point(image, face_location)
face_encodings_list.append(np.array(face_encoder.compute_face_descriptor(image, shape, num_jitters=1)))
shape = face_utils.shape_to_np(shape)
landmarks_list.append(shape)
face_locations = transform(image, face_locations)
return face_encodings_list, face_locations, landmarks_list

def easy_face_reco(frame, known_face_encodings, known_face_names):
rgb_small_frame = frame[:, :, ::-1]
# ENCODING FACE
face_encodings_list, face_locations_list, landmarks_list = encode_face(rgb_small_frame)
face_names = []
for face_encoding in face_encodings_list:
if len(face_encoding) == 0:
return np.empty((0))
# CHECK DISTANCE BETWEEN KNOWN FACES AND FACES DETECTED
vectors = np.linalg.norm(known_face_encodings - face_encoding, axis=1)
tolerance = 0.6
result = []
for vector in vectors:
if vector <= tolerance:
result.append(True)
else:
result.append(False)
if True in result:
first_match_index = result.index(True)
name = known_face_names[first_match_index]

    else:
        name = "Unknown"
    face_names.append(name)

for (top, right, bottom, left), name in zip(face_locations_list, face_names):
         cv2.rectangle(frame, (left, top), (right, bottom), (0, 255, 0), 2)
         cv2.rectangle(frame, (left, bottom - 30), (right, bottom), (0, 255, 0), cv2.FILLED)
         cv2.putText(frame, name, (left + 2, bottom - 2), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 0, 0), 1)

for shape in landmarks_list:
    for (x, y) in shape:
           cv2.circle(frame, (x, y), 1, (255, 0, 255), -1)

if __name__ == '__main__':
print('[INFO] Importing faces...')
face_to_encode_path =['input/ahmed12.png']
known_face_encodings = []
for face_to_encode_path in face_to_encode_path:
image = PIL.Image.open(face_to_encode_path)
image = np.array(image)
face_encoded = encode_face(image)[0][0]
known_face_encodings.append(face_encoded)
known_face_names = ["taleb"]
print('[INFO] starting webcam ...')
video_capture = cv2.VideoCapture(0)
print('[INFO] Webcam well started ')
print('[INFO] Detecting...')
while True:
ret,frame = video_capture.read()
easy_face_reco(frame,known_face_encodings,known_face_names)
cv2.imshow('live Reconnaissance faciale App', frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
print('[INFO] Stopping System')
video_capture.release()
cv2.destroyAllWindows()
il afficher l'erreur suivant;
Traceback (most recent call last):
File "C:/easy_facial_recognition-master/tareq.py", line 79, in
face_encoded = encode_face(image)[0][0]
IndexError: list index out of range