Commit 34f058e3 authored by rico.j's avatar rico.j
Browse files

making the prints even nicer

parent 055c0a84
......@@ -175,7 +175,7 @@ def experiment_APC(datasets, results_path, only_berlin_data=False):
with open(result_file, 'wb') as f:
pickle.dump(results, f, 2)
print('segmentation results are written to {}'.format(path))
print('Segmentation results are written to {}'.format(path))
return result_file
......@@ -950,32 +950,34 @@ if __name__ == "__main__":
datasets['berlin_runs'] = combine_datasets([datasets["berlin_runs/"+str(i+1)] for i in range(3)])
datasets['seattle_runs'] = combine_datasets([datasets["seattle_runs/"+str(i+1)] for i in range(5)])
plt.ion()
#plt.ion()
# vizualize the dataset
#datasets['seattle_test'].visualize_dataset()
print('A. Performance Evaluation')
print('\nThis script repeats all experiments from the paper "Probabilistic Object Segmentation for the Amazon Picking Challenge".')
print('\nIt may take a while ... :D')
print('\nA. Performance Evaluation')
print('1) Performance in the Amazon Picking Challenge')
experiment_APC(datasets, results_path)
print('2) Performance by Object')
print('\n2) Performance by Object')
filename_results = experiment_our_method(datasets, results_path)
display_experiment_objects(filename_results)
print('3) Increasing the Number of Objects per Bin')
print('\n3) Increasing the Number of Objects per Bin')
filename_results = experiment_candidates(datasets, results_path)
display_experiment_candidates(filename_results)
print('B. Comparison to CRF')
print('\nB. Comparison to CRF')
filename_results = experiment_baseline(datasets, results_path)
display_experiment_baseline(filename_results)
print('C. Variants of the Algorithm')
print('\nC. Variants of the Algorithm')
print('1) Changing Features')
filename_result = []
......@@ -983,19 +985,19 @@ if __name__ == "__main__":
filename_result.append(experiment_removed_features(datasets, results_path))
display_experiment_features(filename_result)
print('2) Pixel Labeling and Selection')
print('\n2) Pixel Labeling and Selection')
filename_results = experiment_segmentation(datasets, results_path)
display_experiment_segmentation(filename_results)
print('3) Re-Labeling and Post-Processing')
print('\n3) Re-Labeling and Post-Processing')
filename_results = experiment_reseg_and_postproc(datasets, results_path)
display_experiment_reseg_and_postproc(filename_results)
print('4) Random Forest for Pixel Probability Estimation')
print('\n4) Random Forest for Pixel Probability Estimation')
filename_results = experiment_model(datasets, results_path)
display_experiment_model(filename_results)
print('done')
\ No newline at end of file
print('\nDone.')
\ No newline at end of file
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment