search

facebookresearch / localrf

An algorithm for reconstructing the radiance field of a large-scale scene from a single casually captured video.
MIT License
956 stars 62 forks source link

Results on static hikes #38

Open tb2-sy opened 10 months ago

tb2-sy commented 10 months ago

Hello! Thank you for the nice work! I noticed that the paper only shows the average of five scenes perfmance in the static hikes dataset.Could you please share a table with the results on each scene? I mean for each of the 12 scenes. Thank you very much.

ameuleman commented 8 months ago
Hi, here are results for each scene. All methods are self-calibrated and the table shows the five scenes the ablation (against Ours w/o progressive optimization and Ours w/o local RF) has been conducted on. Method Scene PSNR SSIM LPIPS
barf forest1 12.85 0.244 0.944
forest2 14.53 0.222 0.914
forest3 10.74 0.250 0.964
garden1 11.50 0.194 0.983
garden2 13.80 0.202 1.032
garden3 12.75 0.283 0.914
indoor 14.81 0.692 0.737
university1 12.34 0.382 0.860
university2 11.39 0.294 0.875
university3 11.28 0.281 0.954
university4 12.11 0.464 0.820
playground 12.24 0.201 0.962
Ours w/o progressive optimization forest1 12.81 0.286 0.937
forest2 16.32 0.232 0.894
forest3 14.54 0.265 0.916
playground 14.45 0.226 0.963
university1 15.76 0.386 0.872
Ours w/o local RF forest1 14.17 0.306 0.886
forest2 18.47 0.278 0.792
forest3 14.87 0.268 0.894
playground 16.58 0.260 0.847
university1 18.68 0.448 0.731
ours forest1 16.06 0.368 0.737
forest2 25.08 0.717 0.345
forest3 18.21 0.394 0.660
garden1 16.85 0.340 0.685
garden2 18.67 0.316 0.741
garden3 21.93 0.544 0.483
indoor 27.52 0.849 0.364
university1 22.35 0.674 0.384
university2 21.55 0.646 0.381
university3 19.67 0.581 0.414
university4 22.32 0.657 0.392
playground 18.32 0.351 0.648
tb2-sy commented 8 months ago

Hi, here are results for each scene. All methods are self-calibrated and the table shows the five scenes the ablation (against Ours w/o progressive optimization and Ours w/o local RF) has been conducted on.

Method Scene PSNR SSIM LPIPS barf forest1 12.85 0.244 0.944 forest2 14.53 0.222 0.914 forest3 10.74 0.250 0.964 garden1 11.50 0.194 0.983 garden2 13.80 0.202 1.032 garden3 12.75 0.283 0.914 indoor 14.81 0.692 0.737 university1 12.34 0.382 0.860 university2 11.39 0.294 0.875 university3 11.28 0.281 0.954 university4 12.11 0.464 0.820 playground 12.24 0.201 0.962 Ours w/o progressive optimization forest1 12.81 0.286 0.937 forest2 16.32 0.232 0.894 forest3 14.54 0.265 0.916 playground 14.45 0.226 0.963 university1 15.76 0.386 0.872 Ours w/o local RF forest1 14.17 0.306 0.886 forest2 18.47 0.278 0.792 forest3 14.87 0.268 0.894 playground 16.58 0.260 0.847 university1 18.68 0.448 0.731 ours forest1 16.06 0.368 0.737 forest2 25.08 0.717 0.345 forest3 18.21 0.394 0.660 garden1 16.85 0.340 0.685 garden2 18.67 0.316 0.741 garden3 21.93 0.544 0.483 indoor 27.52 0.849 0.364 university1 22.35 0.674 0.384 university2 21.55 0.646 0.381 university3 19.67 0.581 0.414 university4 22.32 0.657 0.392 playground 18.32 0.351 0.648

Thank you very much for your reply. Could you please tell me which five scenes average performance are reported in the ablation experiment part of your paper?

ameuleman commented 8 months ago

forest1, forest2, forest3, playground, university1

tb2-sy commented 8 months ago

Thank you for such a quick reply!