Pascal VOC output is black

[energabos]

has anyone tried using the weight files for Pascal VOC with the webcam_demo.py?

import numpy as np import matplotlib.pyplot as plt import os.path import scipy import argparse import math import cv2 import sys import time sys.path.append('/usr/local/lib/python2.7/site-packages') caffe_root = '/home/ubuntu/caffe-segnet-cudnn5/' sys.path.insert(0, caffe_root + 'python') import caffe parser = argparse.ArgumentParser() parser.add_argument('--model', type=str, required=True) parser.add_argument('--weights', type=str, required=True) parser.add_argument('--colours', type=str, required=True) args = parser.parse_args() net = caffe.Net(args.model, args.weights, caffe.TEST) caffe.set_mode_gpu() input_shape = net.blobs['data'].data.shape output_shape = net.blobs['prob'].data.shape label_colours = cv2.imread(args.colours).astype(np.uint8) cv2.namedWindow("Input") cv2.namedWindow("SegNet") cap = cv2.VideoCapture("/home/ubuntu/vid.mp4") rval = True while rval: start = time.time() rval, frame = cap.read() if rval == False: break end = time.time() print '%30s' % 'Grabbed camera frame in ', str((end - start)*1000), 'ms' start = time.time() frame = cv2.resize(frame, (input_shape[3],input_shape[2])) input_image = frame.transpose((2,0,1))

input_image = input_image[(2,1,0),:,:] # May be required, if you do not open your data with opencv

input_image = np.asarray([input_image]) end = time.time() print '%30s' % 'Resized image in ', str((end - start)1000), 'ms' start = time.time() out = net.forward_all(data=input_image) end = time.time() print '%30s' % 'Executed SegNet in ', str((end - start)1000), 'ms' start = time.time() segmentation_ind = np.squeeze(net.blobs['prob'].data) segmentation_ind_3ch = np.resize(segmentation_ind,(3,input_shape[2],input_shape[3])) segmentation_ind_3ch = segmentation_ind_3ch.transpose(1,2,0).astype(np.uint8) segmentation_rgb = np.zeros(segmentation_ind_3ch.shape, dtype=np.uint8) cv2.LUT(segmentation_ind_3ch,label_colours,segmentation_rgb) segmentation_rgb = segmentation_rgb.astype(float)/255 end = time.time() print '%30s' % 'Processed results in ', str((end - start)*1000), 'ms\n cv2.imshow("Input", frame) cv2.imshow("SegNet", segmentation_rgb)
key = cv2.waitKey(1) if key == 27: # exit on ESC break cap.release() cv2.destroyAllWindows()

what should i use for the --colours arg here?

Thank you!

[Arsey]

@energabos did you get out = net.forward_all(data=input_image) working? for me it raises an exception Input blob arguments do not match net inputs