for i in range(gmphd_num_steps):
sys.stdout.write('Step %d\r\n' % i)
gm_list = gmphd_predictor(model=gmphd_model, gm_list=gmphd_gm_list)
gm_list = gmphd_corrector(model=gmphd_model,
gm_list=gm_list,
observation=gmphd_observation[i])
gm_list = gm_pruning(gm_list=gm_list,
T=gmphd_model.gm_T,
U=gmphd_model.gm_U,
C=gmphd_model.gm_Jmax)
# Get Prediction
gmphd_prediction.append(gm_estimator(gm_list))
# Save temporary result
gmphd_phd['s'].append(
gm_calculate(gm_list=gm_list, grid=(grid_x, grid_y, grid_z)))
gmphd_phd['gms'].append(copy.deepcopy(gm_list))
# Update GM_LIST
gmphd_gm_list = gm_list
print('Target spotted %d' % len(gmphd_prediction[i]))
result_filename = os.path.join(chkpt_dir, 'gmphd_3dcv_chkpt.pkl')
result_fp = open(result_filename, 'wb')
pickle.dump(
{
'model': gmphd_model,
'phd': gmphd_phd,
'prediction': gmphd_prediction,
sys.stdout.write('Step %d\r\n' % i)
gm_list = gmphd_predictor(model=gmphd_model,
gm_list=gmphd_gm_list)
gm_list = gmphd_corrector(model=gmphd_model,
gm_list=gm_list,
observation=gmphd_observation[i])
gm_list = gm_pruning(gm_list=gm_list,
T=gmphd_model.gm_T,
U=gmphd_model.gm_U,
C=gmphd_model.gm_Jmax)
# Get Prediction
gmphd_prediction.append(gm_estimator(gm_list))
# Save temporary result
gmphd_phd['s'].append(gm_calculate(
gm_list=gm_list,
grid=(grid_x, grid_y)
))
gmphd_phd['gms'] = copy.deepcopy(gm_list)
# Update GM_LIST
gmphd_gm_list = gm_list
print('Target spotted %d' % len(gmphd_prediction[i]))
result_filename = os.path.join(chkpt_dir, 'gmphd_2dcv_noise_chkpt.pkl')
result_fp = open(result_filename, 'wb')
pickle.dump({
'model': gmphd_model,
'phd': gmphd_phd,
'prediction': gmphd_prediction,