Question on results about geotracknet HOT 2 CLOSED

Hello @lynnetab ,

1. "In your paper, "GeoTrackNet A Maritime Anomaly Detector using Probabilistic Neural Network Rep. of AIS Tracks and a Contrario Detection" you report than you found only 25 anomalies. Please let me know if the software has been updated for this paper since I have 26 anomalies."
   => That is not shocking. Different hardware represents floating-point numbers differently, so we cannot guarantee that the results will always be identical. Are your 26 anomalies consistent with the 25 presented in the paper?

2. SPEED_MAX is 30 knots.

3. v_sog should be tmp[:,2]*SPEED_MAX/float(onehot_sog_bins) = tmp[:,2]*30/30.

4. v_cog should be tmp[:,3]*COG_MAX/float(onehot_cog_bins) = tmp[:,3]*360/72.

And yes, please share the results with us and the community.

from geotracknet.

I was able to run the GeoTrackNet software after following the instructions and using the provided datasets. Here are my notes:

• python 3.6.10
• Install relevant python libraries requirements file (conda env update -v --file requirements.yml)
• Ran the code with 1 CPU and 1 GPU.
• Ran the code uses the steps provide on summary page along with the test, training, and validation data sets: ct_2017010203_10_20/ct_2017010203_10_20_test.pkl, ct_2017010203_10_20/ct_2017010203_10_20_train.pkl and ct_2017010203_10_20/ct_2017010203_10_20_valid.pkl

Results:
I had the following 26 MMSI identified as abnormal.
Abnormal_tracks-ct_2017010203_10_20_train.pkl-ct_2017010203_10_20_test.pkl-latent_size-100-step-80003-eps-1e-10-26.csv

from geotracknet.