append loadables resnet

resnet18-imagenet-caffe/predict.py

@@ -0,0 +1,37 @@
+import caffe
+import numpy as np
+from PIL import Image
+def inference(prototxt, caffemodel, image):
+    net = caffe.Net(prototxt,
+            caffemodel,
+            caffe.TEST)
+    mean = np.ones([3, 224, 224], dtype=np.float)
+    mean[0,:,:] = 104
+    mean[1,:,:] = 117
+    mean[2,:,:] = 123
+    # create transformer for the input called 'data'
+    transformer = caffe.io.Transformer({'data': net.blobs['data'].data.shape})
+    transformer.set_transpose('data', (2,0,1))  # move image channels to outermost dimension
+    transformer.set_mean('data', mean)            # subtract the dataset-mean value in each channel
+    transformer.set_raw_scale('data', 255)  # rescale from [0, 1] to [0, 255]
+    transformer.set_input_scale('data', 1.0)
+    # transformer.set_channel_swap('data', (2,1,0))  # swap channels from RGB to BGR
+    image = caffe.io.load_image(image)
+    print(image.shape)
+    transformed_image = transformer.preprocess('data', image)
+    out = net.forward_all(data=np.asarray([transformed_image]))
+    return out
+
+prototxt_path = 'deploy.prototxt'
+caffemodel_path = 'resnet-18.caffemodel'
+img_path = 'raw/aircraft_carrier.jpg'
+labels_file = 'synset_words.txt'
+prediction = inference(prototxt_path, caffemodel_path, img_path)
+labels = np.loadtxt(labels_file, str, delimiter='\t')
+output_prob = prediction['prob'][0]
+# sort top five predictions from softmax output
+top_inds = output_prob.argsort()[::-1][:5]  # reverse sort and take five largest items
+print('probabilities and labels:')
+for i in top_inds:
+    print(i,output_prob[i], labels[i])
+

resnet18-imagenet-caffe/predict_tranformed.py

@@ -0,0 +1,27 @@
+import caffe
+import numpy as np
+import cv2 as cv
+def inference(prototxt, caffemodel, image):
+    net = caffe.Net(prototxt,
+            caffemodel,
+            caffe.TEST)
+    image = cv.imread(image)
+    print(image.shape)
+    image = image.transpose((2,0,1))
+    print(image)
+    out = net.forward_all(data=np.asarray([image]))
+    return out
+
+prototxt_path = 'deploy.prototxt'
+caffemodel_path = 'resnet-18.caffemodel'
+img_path = '0_tench.jpg'
+labels_file = 'synset_words.txt'
+prediction = inference(prototxt_path, caffemodel_path, 'transformed/' + img_path)
+labels = np.loadtxt(labels_file, str, delimiter='\t')
+output_prob = prediction['prob'][0]
+# sort top five predictions from softmax output
+top_inds = output_prob.argsort()[::-1][:5]  # reverse sort and take five largest items
+print('probabilities and labels:')
+for i in top_inds:
+    print(i,output_prob[i], labels[i])
+

resnet18-imagenet-caffe/resize.py

@@ -0,0 +1,49 @@
+import caffe
+import numpy as np
+from PIL import Image
+import cv2 as cv
+def inference(prototxt, caffemodel, image):
+    net = caffe.Net(prototxt,
+            caffemodel,
+            caffe.TEST)
+    mean = np.ones([3, 224, 224], dtype=np.float)
+    mean[0,:,:] = 104
+    mean[1,:,:] = 117
+    mean[2,:,:] = 123
+    # create transformer for the input called 'data'
+    transformer = caffe.io.Transformer({'data': net.blobs['data'].data.shape})
+    transformer.set_transpose('data', (2,0,1))  # move image channels to outermost dimension
+    # transformer.set_mean('data', mean)            # subtract the dataset-mean value in each channel
+    transformer.set_raw_scale('data', 255)  # rescale from [0, 1] to [0, 255]
+    transformer.set_channel_swap('data', (2,1,0))  # swap channels from RGB to BGR
+    image = caffe.io.load_image(image)
+    transformed_image = transformer.preprocess('data', image)
+    print(transformed_image)
+    out = net.forward_all(data=np.asarray([transformed_image]))
+    return transformed_image,out
+
+prototxt_path = 'deploy.prototxt'
+caffemodel_path = 'resnet-18.caffemodel'
+img_path = 'aircraft_carrier.jpg'
+labels_file = 'synset_words.txt'
+transformed_image, prediction = inference(prototxt_path, caffemodel_path, 'raw/'+img_path)
+labels = np.loadtxt(labels_file, str, delimiter='\t')
+output_prob = prediction['prob'][0]
+# sort top five predictions from softmax output
+top_inds = output_prob.argsort()[::-1][:5]  # reverse sort and take five largest items
+print('probabilities and labels:')
+for i in top_inds:
+    print(i,output_prob[i], labels[i])
+
+dirname = 'resized/'
+filepath = dirname + str(int(top_inds[0])) + '_' + img_path
+print(filepath)
+saveim = transformed_image.transpose((1, 2, 0))
+cv.imwrite(filepath, saveim)
+# saveim = Image.fromarray(saveim.astype(np.uint8))
+# print('-----------------')
+# print(np.array(saveim))
+# saveim.save(filepath)
+# loadim = Image.open(filepath)
+# print('+++++++++++++++')
+# print(np.array(loadim))