I'm trying to infer the model on DIODE dataset to reproduce the result of Table 2 in the paper. However, as there are many noisy points in the ground truth, and also I don't know the exact mapping relationship between the output disparity and relative depth, the metrics are very bad. Could you please release your test code on DIODE dataset?
I'm trying to infer the model on DIODE dataset to reproduce the result of Table 2 in the paper. However, as there are many noisy points in the ground truth, and also I don't know the exact mapping relationship between the output disparity and relative depth, the metrics are very bad. Could you please release your test code on DIODE dataset?