search

sconlyshootery / FeatDepth

This is the offical codes for the methods described in the "Feature-metric Loss for Self-supervised Learning of Depth and Egomotion".
MIT License
247 stars 28 forks source link

Is there a way to visually check the output of the FeatureNet #74

Closed Kletterer closed 3 years ago

Kletterer commented 3 years ago

Hi, Thanks for the great paper! I am interested in the FeatureNet as I am writing my master thesis about monocular depth estimation.

Testing the depth-map works great using the infer.py script, but is there a way to save the output of the encoder part of the FeatureNet? Similar to Fig. 3 in the paper?

Warm regards!

sconlyshootery commented 3 years ago

You can use PCA decomposition to find 3 channels for visualization as a rgb image, you can also save the maximum along the channels to get one channel image.