Underwater Visual Localization Using Machine Learning and LSTM: Experiments, and References | HackerNoon
Briefly

from Hackernoon8 months ago
We test three model configurations on both the simulator and tank datasets, showing that ResNet-50 and LSTM models perform well in simulated settings but vary in performance in tank datasets with typical underwater distortions.
Hackernoonhttps://hackernoon.com/underwater-visual-localization-using-machine-learning-and-lstm-experiments-and-references
With the simulator dataset without challenges present in real underwater scenarios, ResNet-50 and LSTM-based models outperformed the baseline model in predicting trajectories accurately.
Hackernoonhttps://hackernoon.com/underwater-visual-localization-using-machine-learning-and-lstm-experiments-and-references
Read at Hackernoon
#underwater-navigation #model-configurations #resnet-50 #lstm-models #data-analysis
[
|
]