def main(args):
rs=[]
for i in range(1,repeat):
print('doing{}'.format(i))
test_loader = custom_data_loader.benchmarkLoader(args)
model = custom_model.buildModel(args)
# print(sum(p.numel() for p in model.parameters() if p.requires_grad))
recorder = recorders.Records(args.log_dir)
# r=test_utils.test_split2c(args, 'test', test_loader, model, log, 1, recorder, padding=4, stride=128)
r=test_utils.test_split_rob(args, 'test', test_loader, model, log, 1, recorder, padding=16, stride=96, noise_level=i)
# r=test_utils.test(args, 'test', test_loader, model, log, 1, recorder)
r.append(np.mean(r))
rs.append(np.stack(r))
# test_utils.test(args, 'test', test_loader, model, log, 1, recorder)
# rs=np.stack(rs)
# every_mean=np.mean(rs,axis=0)
# mean=np.mean(rs)
# print(every_mean,mean)
print(rs)
# rs=np.stack(rs,axis=0)
# rs=np.mean(rs,axis=0)
workbook = xlwt.Workbook(encoding='ascii')
worksheet = workbook.add_sheet('My Worksheet', cell_overwrite_ok=True)
for i,r in zip(range(repeat),rs):
for ii,rr in zip(range(10),r):
worksheet.write(i + 1, ii, label=rr)
workbook.save('result/robust/pro.xls')
def main(args):
rs = []
for i in range(15, 16):
# path="data/Training5shadow/checkp_{}.pth.tar".format(i)
# args.retrain=path
log = logger.Logger(args)
test_loader = custom_data_loader.benchmarkLoader(args)
model = custom_model.buildModel(args)
# print(sum(p.numel() for p in model.parameters() if p.requires_grad))
recorder = recorders.Records(args.log_dir)
# r=test_utils.test_split2c(args, 'test', test_loader, model, log, 1, recorder, padding=4, stride=128)
r = test_utils.test_split(args,
'test',
test_loader,
model,
log,
1,
recorder,
padding=0,
stride=64)
rs.append(np.mean(r))
# test_utils.test(args, 'test', test_loader, model, log, 1, recorder)
# rs=np.stack(rs)
# every_mean=np.mean(rs,axis=0)
# mean=np.mean(rs)
# print(every_mean,mean)
print(rs)
f = open('runlog.txt', 'w+')
f.writelines(np.array2string(rs))
f.close()
def main(args):
rs = []
for i in range(repeat):
args.in_img_num = -4 + i
# path="data/Training5shadow/checkp_{}.pth.tar".format(i)
# args.retrain=path
log = logger.Logger(args)
test_loader = custom_data_loader.benchmarkLoader(args)
model = custom_model.buildModel(args)
# print(sum(p.numel() for p in model.parameters() if p.requires_grad))
recorder = recorders.Records(args.log_dir)
# r=test_utils.test_split2c(args, 'test', test_loader, model, log, 1, recorder, padding=4, stride=128)
r = test_utils.test_split(args,
'test',
test_loader,
model,
log,
1,
recorder,
padding=10,
stride=120)
rs.append(r)
# test_utils.test(args, 'test', test_loader, model, log, 1, recorder)
# rs=np.stack(rs)
# every_mean=np.mean(rs,axis=0)
# mean=np.mean(rs)
# print(every_mean,mean)
print(rs)
workbook = xlwt.Workbook(encoding='ascii')
worksheet = workbook.add_sheet('My Worksheet', cell_overwrite_ok=True)
for i, r in zip(range(repeat), rs):
for ii, rr in zip(range(9), r):
worksheet.write(i + 1, ii, label=rr)
workbook.save('result/sparse/sparse_pro{}.xls'.format(i))
def main(args):
model = custom_model.buildModelStage0(args)
val_loader = custom_data_loader.benchmarkLoader(args)
#test_loader = custom_data_loader.benchmarkLoader(args)
#model = custom_model.buildModel(args)
recorder = recorders.Records(args.log_dir)
test_utils.test(args, 'val', val_loader, model, log, 1, recorder)
log.plotCurves(recorder, 'val')
def main(args):
test_loader = custom_data_loader.benchmarkLoader(args)
model = custom_model.buildModel(args)
model_s2 = custom_model.buildModelStage2(args)
models = [model, model_s2]
recorder = recorders.Records(args.log_dir)
test_utils.test(args, 'test', test_loader, models, log, 1, recorder)
def main(args):
model = custom_model.buildModelStage3(args)
recorder = recorders.Records(args.log_dir)
val_loader = custom_data_loader.benchmarkLoader(args)
#train_loader, val_loader = custom_data_loader.reflectanceDataloader(args)
test_utils.testOnBm(args, 'val', val_loader, model, log, 1, recorder)
log.plotCurves(recorder, 'val')
def main(args):
rs=[]
for i in range(5,6):
# path="data/Training5shadow/checkp_{}.pth.tar".format(i)
# args.retrain=path
log = logger.Logger(args)
test_loader = custom_data_loader.benchmarkLoader(args)
model = custom_model.buildModel(args)
# print(sum(p.numel() for p in model.parameters() if p.requires_grad))
recorder = recorders.Records(args.log_dir)
# r=test_utils.test_split2c(args, 'test', test_loader, model, log, 1, recorder, padding=4, stride=128)
test_utils.estimate(args, i, test_loader, model, log, 1, recorder, padding=4, stride=64, split=True)
def main(args):
rs = []
# path="data/Training4shadow/checkp_{}.pth.tar".format(i)
# args.retrain=path
test_loader = custom_data_loader.benchmarkLoader(args)
model = custom_model.buildModel(args)
recorder = recorders.Records(args.log_dir)
# r=test_utils.test_split(args, 'test', test_loader, model, log, 1, recorder, padding=4, stride=160)
# errors=test_utils.test(args, 'test', test_loader, model, log, 1, recorder)
errors = test_utils.test2c(args, 'test', test_loader, model, log, 1,
recorder)
csvfile = open('names.csv', 'w', newline='')
obj = csv.writer(csvfile)
for person in errors:
obj.writerow(person)
csvfile.close()
def main(args):
test_loader = custom_data_loader.benchmarkLoader(args)
model = custom_model.buildModel(args)
recorder = recorders.Records(args.log_dir)
test_utils.test(args, 'test', test_loader, model, log, 1, recorder)
log.plotCurves(recorder, 'test')