VGG Face Descriptor in python with caffe

Question

I want implement VGG Face Descriptor in python. But I keep getting an error:

TypeError: can only concatenate list (not "numpy.ndarray") to list

My code:

import numpy as np
import cv2 
import caffe
img = cv2.imread("ak.png")
img = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
net = caffe.Net("VGG_FACE_deploy.prototxt","VGG_FACE.caffemodel",  caffe.TEST)
print net.forward(img)

Can you help me ?

UPDATE 1

This working code is example in matlab

%  Copyright (c) 2015, Omkar M. Parkhi
%  All rights reserved.
img = imread('ak.png');
img = single(img);

    Img = [129.1863,104.7624,93.5940] ;

img = cat(3,img(:,:,1)-averageImage(1),...
    img(:,:,2)-averageImage(2),...
    img(:,:,3)-averageImage(3));

img = img(:, :, [3, 2, 1]); % convert from RGB to BGR
img = permute(img, [2, 1, 3]); % permute width and height

model = 'VGG_FACE_16_deploy.prototxt';
weights = 'VGG_FACE.caffemodel';
caffe.set_mode_cpu();
net = caffe.Net(model, weights, 'test'); % create net and load weights

res = net.forward({img});
prob = res{1};

caffe_ft = net.blobs('fc7').get_data();

Corticosterone

why aren't you using caffe.io.load_image? – Sooksoon

If i try caffe.io.load_image s i get same error TypeError: can only concatenate list (not "numpy.ndarray") to list. If I try passing a single element list to method i get error TypeError: unhashable type: 'numpy.ndarray' – Corticosterone

try net.forward_all instead of forward. – Sooksoon

same error TypeError: unhashable type: 'numpy.ndarray' – Corticosterone

can you get a stack trace to show what line of code exactly causes this error? – Sooksoon

The error is caused on line 173 in this file - link and this is the image link – Corticosterone

line 173 is a comment. can you please copy the line? – Sooksoon

sorry. all_outs = {out: [] for out in set(self.outputs + (blobs or []))} . Line 174. – Corticosterone

can you post 'VGG_FACE_16_deploy.prototxt' please? – Sooksoon

There is VGG_FACE_16_deploy.prototxt - link – Corticosterone

can you please try net.forward_all(data=[img])? – Sooksoon

Ok. This is looking better. When I try net.forward_all(data = [img]) I get error AttributeError: 'list' object has no attribute 'shape' (caused by line 101). But when i try net.forward_all(data = img) i get error ValueError: could not broadcast input array from shape (1,224,3) into shape (1,3,224,224) (caused by line 176 and then 103). This maybe resolve UPDATE 1 in my post but i dont know synatx of matlab language. Do you know what happen with image from read image to call res = net.forward({img}); – Corticosterone

Answer 1

To use python interface you need to transform the input image before feeding it to the net

img = caffe.io.load_image( "ak.png" )
img = img[:,:,::-1]*255.0 # convert RGB->BGR
avg = np.array([93.5940, 104.7624, 129.1863])  # BGR mean values
img = img - avg # subtract mean (numpy takes care of dimensions :)

Now img is H-by-W-by-3 numpy array.
Caffe expects its inputs as 4D: batch_index x channel x width x height.
Therefore you need to transpose the input and add a singleton dimension to represent the "batch_index" leading dimension

img = img.transpose((2,0,1)) 
img = img[None,:] # add singleton dimension

Now you can run the forward pass

out = net.forward_all( data = img )

Sooksoon

where does this average come from? – Tippets

@maxou it was part of the question. – Sooksoon

Answer 2

OpenCV reads in BGR and scaled to 255 format by default, so:

img = cv2.imread('ak.png')
avg = np.array([93.5940,104.7624,129.1863]) # BGR mean from VGG
img -= avg # subtract mean
img = img.transpose((2,0,1)) # to match image input dimension: 3x224x224
img = img[None,:] # add singleton dimension to match batch dimension
out = net.forward_all(data = img)

Egis

Answer 3

Try passing a single element list to the method.

net.forward ([img])

Efrainefram

Whetn i try it i get this error: TypeError: unhashable type: 'numpy.ndarray' – Corticosterone