File size: 6,214 Bytes
d4b77ac |
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 |
# --------------------------------------------------------
# Python Single Object Tracking Evaluation
# Licensed under The MIT License [see LICENSE for details]
# Written by Fangyi Zhang
# @author [email protected]
# @project https://github.com/StrangerZhang/pysot-toolkit.git
# Revised for SiamMask by foolwood
# --------------------------------------------------------
import warnings
import itertools
import numpy as np
from colorama import Style, Fore
from ..utils import calculate_failures, calculate_accuracy
class AccuracyRobustnessBenchmark:
"""
Args:
dataset:
burnin:
"""
def __init__(self, dataset, burnin=10):
self.dataset = dataset
self.burnin = burnin
def eval(self, eval_trackers=None):
"""
Args:
eval_tags: list of tag
eval_trackers: list of tracker name
Returns:
ret: dict of results
"""
if eval_trackers is None:
eval_trackers = self.dataset.tracker_names
if isinstance(eval_trackers, str):
eval_trackers = [eval_trackers]
result = {}
for tracker_name in eval_trackers:
accuracy, failures = self._calculate_accuracy_robustness(tracker_name)
result[tracker_name] = {'overlaps': accuracy,
'failures': failures}
return result
def show_result(self, result, eao_result=None, show_video_level=False, helight_threshold=0.5):
"""pretty print result
Args:
result: returned dict from function eval
"""
tracker_name_len = max((max([len(x) for x in result.keys()])+2), 12)
if eao_result is not None:
header = "|{:^"+str(tracker_name_len)+"}|{:^10}|{:^12}|{:^13}|{:^7}|"
header = header.format('Tracker Name',
'Accuracy', 'Robustness', 'Lost Number', 'EAO')
formatter = "|{:^"+str(tracker_name_len)+"}|{:^10.3f}|{:^12.3f}|{:^13.1f}|{:^7.3f}|"
else:
header = "|{:^"+str(tracker_name_len)+"}|{:^10}|{:^12}|{:^13}|"
header = header.format('Tracker Name',
'Accuracy', 'Robustness', 'Lost Number')
formatter = "|{:^"+str(tracker_name_len)+"}|{:^10.3f}|{:^12.3f}|{:^13.1f}|"
bar = '-'*len(header)
print(bar)
print(header)
print(bar)
if eao_result is not None:
tracker_eao = sorted(eao_result.items(),
key=lambda x:x[1]['all'],
reverse=True)[:20]
tracker_names = [x[0] for x in tracker_eao]
else:
tracker_names = list(result.keys())
for tracker_name in tracker_names:
ret = result[tracker_name]
overlaps = list(itertools.chain(*ret['overlaps'].values()))
accuracy = np.nanmean(overlaps)
length = sum([len(x) for x in ret['overlaps'].values()])
failures = list(ret['failures'].values())
lost_number = np.mean(np.sum(failures, axis=0))
robustness = np.mean(np.sum(np.array(failures), axis=0) / length) * 100
if eao_result is None:
print(formatter.format(tracker_name, accuracy, robustness, lost_number))
else:
print(formatter.format(tracker_name, accuracy, robustness, lost_number, eao_result[tracker_name]['all']))
print(bar)
if show_video_level and len(result) < 10:
print('\n\n')
header1 = "|{:^14}|".format("Tracker name")
header2 = "|{:^14}|".format("Video name")
for tracker_name in result.keys():
header1 += ("{:^17}|").format(tracker_name)
header2 += "{:^8}|{:^8}|".format("Acc", "LN")
print('-'*len(header1))
print(header1)
print('-'*len(header1))
print(header2)
print('-'*len(header1))
videos = list(result[tracker_name]['overlaps'].keys())
for video in videos:
row = "|{:^14}|".format(video)
for tracker_name in result.keys():
overlaps = result[tracker_name]['overlaps'][video]
accuracy = np.nanmean(overlaps)
failures = result[tracker_name]['failures'][video]
lost_number = np.mean(failures)
accuracy_str = "{:^8.3f}".format(accuracy)
if accuracy < helight_threshold:
row += f'{Fore.RED}{accuracy_str}{Style.RESET_ALL}|'
else:
row += accuracy_str+'|'
lost_num_str = "{:^8.3f}".format(lost_number)
if lost_number > 0:
row += f'{Fore.RED}{lost_num_str}{Style.RESET_ALL}|'
else:
row += lost_num_str+'|'
print(row)
print('-'*len(header1))
def _calculate_accuracy_robustness(self, tracker_name):
overlaps = {}
failures = {}
all_length = {}
for i in range(len(self.dataset)):
video = self.dataset[i]
gt_traj = video.gt_traj
if tracker_name not in video.pred_trajs:
tracker_trajs = video.load_tracker(self.dataset.tracker_path, tracker_name, False)
else:
tracker_trajs = video.pred_trajs[tracker_name]
overlaps_group = []
num_failures_group = []
for tracker_traj in tracker_trajs:
num_failures = calculate_failures(tracker_traj)[0]
overlaps_ = calculate_accuracy(tracker_traj, gt_traj,
burnin=10, bound=(video.width, video.height))[1]
overlaps_group.append(overlaps_)
num_failures_group.append(num_failures)
with warnings.catch_warnings():
warnings.simplefilter("ignore", category=RuntimeWarning)
overlaps[video.name] = np.nanmean(overlaps_group, axis=0).tolist()
failures[video.name] = num_failures_group
return overlaps, failures
|