performance for different seq_lengths

yzcjtr / GeoNet

Code for GeoNet: Unsupervised Learning of Dense Depth, Optical Flow and Camera Pose (CVPR 2018)

MIT License

723 stars 181 forks source link

performance for different seq_lengths #31

Closed ethliup closed 5 years ago

ethliup commented 5 years ago

Hi, could I ask if you tried to use seq_length=2 to train the depth & pose network, which is the minimal case for SfM? Does it work well? Thank you.

yzcjtr commented 5 years ago

Sorry I didn't experiment with 2 frames setting. But I guess the performance will get worse for the pose network has less input information. Welcome to share your findings.