def main(_):
utils.set_gpus_to_use()
try:
import tensorvision.train
import tensorflow_fcn.utils
except ImportError:
logging.error("Could not import the submodules.")
logging.error("Please execute:"
"'git submodule update --init --recursive'")
exit(1)
with open(tf.app.flags.FLAGS.hypes, 'r') as f:
logging.info("f: %s", f)
hypes = json.load(f)
utils.load_plugins()
runs_dir = 'RUNS'
utils.set_dirs(hypes, tf.app.flags.FLAGS.hypes)
utils._add_paths_to_sys(hypes)
logging.info("Evaluating on Validation data.")
logdir = os.path.join(runs_dir, FLAGS.RUN)
# logging.info("Output images will be saved to {}".format)
ana.do_analyze(logdir)
logging.info("Creating output on test data.")
kitti_test.do_inference(logdir)
logging.info("Analysis for pretrained model complete.")
logging.info("For evaluating your own models I recommend using:"
"`tv-analyze --logdir /path/to/run`.")
logging.info("tv-analysis has a much cleaner interface.")
def main(_):
utils.set_gpus_to_use()
try:
import tensorvision.train
import tensorflow_fcn.utils
except ImportError:
logging.error("Could not import the submodules.")
logging.error("Please execute:"
"'git submodule update --init --recursive'")
exit(1)
with open(tf.app.flags.FLAGS.hypes, 'r') as f:
logging.info("f: %s", f)
hypes = json.load(f)
utils.load_plugins()
if 'TV_DIR_RUNS' in os.environ:
runs_dir = os.path.join(os.environ['TV_DIR_RUNS'],
'KittiSeg')
else:
runs_dir = 'RUNS'
utils.set_dirs(hypes, tf.app.flags.FLAGS.hypes)
utils._add_paths_to_sys(hypes)
train.maybe_download_and_extract(hypes)
maybe_download_and_extract(runs_dir)
logging.info("Evaluating on Validation data.")
logdir = os.path.join(runs_dir, FLAGS.RUN)
# logging.info("Output images will be saved to {}".format)
ana.do_analyze(logdir)
logging.info("Creating output on test data.")
kitti_test.do_inference(logdir)
logging.info("Analysis for pretrained model complete.")
logging.info("For evaluating your own models I recommend using:"
"`tv-analyze --logdir /path/to/run`.")
logging.info("tv-analysis has a much cleaner interface.")
def main(_):
utils.set_gpus_to_use()
try:
import tensorvision.train
import tensorflow_fcn.utils
except ImportError:
logging.error("Could not import the submodules.")
logging.error("Please execute:"
"'git submodule update --init --recursive'")
exit(1)
with open(tf.app.flags.FLAGS.hypes, 'r') as f:
logging.info("f: %s", f)
hypes = json.load(f)
utils.load_plugins()
if 'TV_DIR_RUNS' in os.environ:
runs_dir = os.path.join(os.environ['TV_DIR_RUNS'], 'KittiBox')
else:
runs_dir = 'RUNS'
utils.set_dirs(hypes, tf.app.flags.FLAGS.hypes)
utils._add_paths_to_sys(hypes)
train.maybe_download_and_extract(hypes)
maybe_download_and_extract(runs_dir)
logging.info("Evaluating on Validation data.")
logdir = os.path.join(runs_dir, FLAGS.RUN)
# logging.info("Output images will be saved to {}".format)
ana.do_analyze(logdir, base_path='hypes')
logging.info("Analysis for pretrained model complete.")
logging.info("For evaluating your own models I recommend using:"
"`tv-analyze --logdir /path/to/run`.")
logging.info("")
logging.info(
"Output images can be found in {}/analyse/images.".format(logdir))
def main(_):
utils.set_gpus_to_use()
try:
import tensorvision.train
import tensorflow_fcn.utils
except ImportError:
logging.error("Could not import the submodules.")
logging.error("Please execute:"
"'git submodule update --init --recursive'")
exit(1)
hypes_path = FLAGS.logdir
hypes_path = os.path.join(hypes_path, "model_files/hypes.json")
with open(hypes_path, 'r') as f:
logging.info("f: %s", f)
hypes = json.load(f)
utils.load_plugins()
if 'TV_DIR_RUNS' in os.environ:
runs_dir = os.path.join(os.environ['TV_DIR_RUNS'], 'FacadeSeg')
else:
runs_dir = 'RUNS'
utils.set_dirs(hypes, FLAGS.hypes)
utils._add_paths_to_sys(hypes)
logging.info("Evaluating on Validation data.")
ana.do_analyze(FLAGS.logdir)
logging.info("Segmenting and test data. Creating output.")
ana.do_inference(FLAGS.logdir)
logging.info("Analysis for pretrained model complete.")
def do_inference(hypes, modules, logdir):
"""
Analyze a trained model.
This will load model files and weights found in logdir and run a basic
analysis.
Paramters
---------
logdir : string
folder with logs
"""
data_input, arch, objective, solver = modules
data_dir = hypes['dirs']['data_dir']
if 'TV_DIR_DATA' in os.environ:
data_dir = os.environ['TV_DIR_DATA']
hypes['dirs']['data_dir'] = data_dir
hypes['dirs']['output_dir'] = logdir
# Tell TensorFlow that the model will be built into the default Graph.
with tf.Graph().as_default():
image_pl, label_pl = _create_input_placeholder()
image = tf.expand_dims(image_pl, 0)
if 'whitening' not in hypes['arch'] or \
hypes['arch']['whitening']:
image = tf.image.per_image_whitening(image)
logging.info('Whitening is enabled.')
else:
logging.info('Whitening is disabled.')
# build the graph based on the loaded modules
softmax = build_inference_graph(hypes,
modules,
image=image,
label=label_pl)
# prepaire the tv session
sess_coll = core.start_tv_session(hypes)
sess, saver, summary_op, summary_writer, coord, threads = sess_coll
_load_weights(logdir, sess, saver)
_prepare_output_folder(hypes, logdir)
val_json = os.path.join(hypes['dirs']['eval_out'], 'val.json')
if FLAGS.inspect:
if not os.path.exists(val_json):
logging.error("File does not exist: %s", val_json)
logging.error("Please run kitti_eval in normal mode first.")
exit(1)
else:
with open(val_json, 'r') as f:
eval_dict = json.load(f)
logging.debug(eval_dict)
from IPython import embed
embed()
exit(0)
logging.info("Doing evaluation with Validation Data")
val_file = os.path.join(hypes['dirs']['data_dir'],
hypes['data']['val_file'])
eval_dict = eval_dataset(hypes, val_file, True, sess, image_pl, softmax)
with open(val_json, 'w') as outfile:
# json.dump(eval_dict, outfile, indent=2)
logging.info("Dumping currently not supported")
logging.info("Succesfully evaluated Dataset. Output is written to %s",
val_json)
logging_file = os.path.join(hypes['dirs']['eval_out'], 'eval.log')
filewriter = _get_filewrite_handler(logging_file)
rootlog = logging.getLogger('')
rootlog.addHandler(filewriter)
logging.info('Statistics on Validation Data.')
logging.info('MaxF1 : %4.2f', 100 * eval_dict['MaxF'])
logging.info('BestThresh : %4.2f', 100 * eval_dict['BestThresh'])
logging.info('Avg Precision : %4.2f', 100 * eval_dict['AvgPrec'])
logging.info('')
ind5 = np.where(eval_dict['thresh'] >= 0.5)[0][0]
logging.info('Precision @ 0.5: %4.2f', 100 * eval_dict['precision'][ind5])
logging.info('Recall @ 0.5: %4.2f', 100 * eval_dict['recall'][ind5])
logging.info('TPR @ 0.5: %4.2f', 100 * eval_dict['recall'][ind5])
logging.info('TNR @ 0.5: %4.2f', 100 * eval_dict['TNR'][ind5])
if FLAGS.kitti_eval:
do_kitti_eval_with_training_data(hypes, sess, image_pl, softmax)
rootlog.removeHandler(filewriter)
ana.do_analyze(FLAGS.logdir)