TensorBox/evaluate.py at master · ArnholdInstitute/TensorBox · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
#!/usr/bin/env python

import tensorflow as tf, os, json, subprocess, cv2, argparse, pdb, rtree, numpy as np, pandas, time
from scipy.misc import imread, imresize
from scipy import misc
from train import build_forward
from utils.annolist import AnnotationLib as al
from utils.train_utils import add_rectangles, rescale_boxes
from shapely.geometry import MultiPolygon, box

def get_image_dir(args):
    weights_iteration = int(args.weights.split('-')[-1])
    expname = '_' + args.expname if args.expname else ''
    image_dir = '%s/images_%s_%d%s' % (os.path.dirname(args.weights), os.path.basename(args.test_boxes)[:-5], weights_iteration, expname)
    return image_dir

def get_metrics(gt_boxes, pred_boxes):
    false_positives = 0
    true_positives = 0
    false_negatives = 0
    total_overlap = 0.0

    # Create the RTree out of the ground truth boxes
    idx = rtree.index.Index()
    for j, rect in enumerate(gt_boxes):
        idx.insert(j, tuple(rect))

    gt_mp = MultiPolygon([box(*b) for b in gt_boxes])
    pred_mp = MultiPolygon([box(*b) for b in pred_boxes])

    for i, rect in enumerate(pred_boxes):
        best_jaccard = 0.0
        best_idx = None
        best_overlap = 0.0
        for gt_idx in idx.intersection(rect[:4]):
            gt = gt_boxes[gt_idx]
            intersection = (min(rect[2], gt[2]) - max(rect[0], gt[0])) * (min(rect[3], gt[3]) - max(rect[1], gt[1]))
            rect_area = (rect[2] - rect[0]) * (rect[3] - rect[1])
            gt_area = (gt[2] - gt[0]) * (gt[3] - gt[1])
            union = rect_area + gt_area - intersection
            jaccard = float(intersection) / float(union)
            if jaccard > best_jaccard:
                best_jaccard = jaccard
                best_idx = gt_idx
            if intersection > best_overlap:
                best_overlap = intersection
        if best_idx is None or best_jaccard <= 0.00000000000001:
            false_positives += 1
        else:
            idx.delete(best_idx, gt_boxes[best_idx])
            true_positives += 1
        total_overlap = best_overlap
    total_jaccard = total_overlap / (gt_mp.area + pred_mp.area - total_overlap) if len(gt_boxes) > 0 else None
    false_negatives = len(gt_boxes) - true_positives

    return false_positives, false_negatives, true_positives, total_jaccard


def get_results(args, H):
    tf.reset_default_graph()
    H["grid_width"] = H["image_width"] / H["region_size"]
    H["grid_height"] = H["image_height"] / H["region_size"]
    x_in = tf.placeholder(tf.float32, name='x_in', shape=[H['image_height'], H['image_width'], 3])
    if H['use_rezoom']:
        pred_boxes, pred_logits, pred_confidences, pred_confs_deltas, pred_boxes_deltas = build_forward(H, tf.expand_dims(x_in, 0), 'test', reuse=None)
        grid_area = H['grid_height'] * H['grid_width']
        pred_confidences = tf.reshape(tf.nn.softmax(tf.reshape(pred_confs_deltas, [grid_area * H['rnn_len'], 2])), [grid_area, H['rnn_len'], 2])
        if H['reregress']:
            pred_boxes = pred_boxes + pred_boxes_deltas
    else:
        pred_boxes, pred_logits, pred_confidences = build_forward(H, tf.expand_dims(x_in, 0), 'test', reuse=None)
    saver = tf.train.Saver()
    all_preditions = []
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())

        saver.restore(sess, args.weights)

        pred_annolist = al.AnnoList()

        true_annolist = al.parse(args.test_boxes)
        data_dir = os.path.join(os.path.dirname(args.test_boxes))

        false_positives, false_negatives, true_positives = 0,0,0


        total_time = 0.0

        image_dir = get_image_dir(args)
        subprocess.call('mkdir -p %s' % image_dir, shell=True)
        for i in range(len(true_annolist)):
            true_anno = true_annolist[i]

            orig_img = imread('%s/%s' % (data_dir, true_anno.imageName))[:,:,:3]
            img = imresize(orig_img, (H["image_height"], H["image_width"]), interp='cubic')
            feed = {x_in: img}

            t0 = time.time()
            (np_pred_boxes, np_pred_confidences) = sess.run([pred_boxes, pred_confidences], feed_dict=feed)
            total_time += time.time() - t0

            pred_anno = al.Annotation()
            pred_anno.imageName = true_anno.imageName
            new_img, rects, all_rects = add_rectangles(H, [img], np_pred_confidences, np_pred_boxes,
                                            use_stitching=True, rnn_len=H['rnn_len'], min_conf=args.min_conf, tau=args.tau, show_suppressed=args.show_suppressed)
            pred_anno.rects = rects
            pred_anno.imagePath = os.path.abspath(data_dir)
            pred_anno = rescale_boxes((H["image_height"], H["image_width"]), pred_anno, orig_img.shape[0], orig_img.shape[1])
            pred_annolist.append(pred_anno)

            all_preditions.extend([[r.x1, r.y1, r.x2, r.y2, r.score, i] for r in all_rects])

            prediction = np.array([[r.x1, r.y1, r.x2, r.y2, r.score] for r in rects])
            targets = np.array([[r.x1, r.y1, r.x2, r.y2] for r in true_anno.rects])

            fp, fn, tp, jaccard = get_metrics(targets, prediction)
            false_positives += fp
            false_negatives += fn
            true_positives += tp

            precision = np.float64(true_positives)/(true_positives + false_positives)
            recall = np.float64(true_positives)/(true_positives + false_negatives)

            print('[%d/%d]: False positives: %d, False negatives: %d, True positives: %d, Precision: %f, Recall: %f' %
                (i, len(true_annolist), false_positives, false_negatives, true_positives, precision, recall))

    df = pandas.DataFrame(all_preditions)
    df.columns = ['x1', 'y1', 'x2', 'y2', 'score', 'image_id']

    print('Total time: %.4f seconds, per image: %.4f' % (total_time, total_time / len(true_annolist)))

    return df

def main():
    parser = argparse.ArgumentParser()

    parser.add_argument('--weights', help='Path to weight file to be used by the model', required=True)
    parser.add_argument('--expname', default='')

    parser.add_argument('--test_boxes', required=True, help='Path to the JSON file containing the test set')
    parser.add_argument('--gpu', default=0)
    parser.add_argument('--logdir', default='output')
    parser.add_argument('--iou_threshold', default=0.5, type=float)
    parser.add_argument('--tau', default=0.25, type=float)
    parser.add_argument('--min_conf', default=0.2, type=float)
    parser.add_argument('--show_suppressed', default=False, type=bool)
    args = parser.parse_args()

    os.environ['CUDA_VISIBLE_DEVICES'] = str(args.gpu)
    hypes_file = '%s/hypes.json' % os.path.dirname(args.weights)
    with open(hypes_file, 'r') as f:
        H = json.load(f)
    expname = args.expname + '_' if args.expname else ''
    pred_boxes = '%s.%s%s' % (args.weights, expname, os.path.basename(args.test_boxes))
    true_boxes = '%s.gt_%s%s' % (args.weights, expname, os.path.basename(args.test_boxes))

    df = get_results(args, H)
    df.to_csv('predictions.csv', index=False)

if __name__ == '__main__':
    main()