Open jeethesh-pai opened 2 years ago
same question! Have you sloved?
No, It needs a good weighting factor for combined loss. The weighting factor given in this repo and mentioned in the paper is not working for me. For the both factors my loss is diverging.
Hallo @rpautrat,
I am very grateful for your repository. It helped me a lot to understand the implementation of the paper Superpoint. I am currently doing a student project on the performance of superpoint on laser-scanned images. Since the tensorflow version is not compatible with the server I am deploying, which I have limited permissions I had to adapt the code for the newer version. I followed your style of code and was successful till training magicpoint for 2 Homographic adaptations. But the descriptor training is not happening. The validation loss diverges from the starting of iteration. I have some questions regarding descriptor training and since you were succeded in training it, i think only you can help me with this doubt.
coord_cells
-warped_coord_cells
should be less than 8. Supposing we use a homography with just translationt_x
to right side with 10 pixels. Applying this transformation will shift thewarped_coord_cells
to 10 pixel right of originalcoords
. That means in the correspondence matrix s(hwh'w'), the first cell is not matching with the first cell ofwarped_coord_cell
whereas it would match with second cell ofwarped_coord_cell
is that right?desc_product
will help training better or do you think removing the both normalization oflast layer of descriptor convolution
and normalization ofdesc_product
helps training better.Training info: I used learning rate of 0.01 as well as 0.001 with Adam and homographic adaptation config
for the descriptor training.
Thank you for the help