What are `lambda_local` and `lambda_d` before $\eta$ and $\Delta_k$ respectively?

[Pirate-Pianist]

Hello. In paper 3.3 Dymamic Range part, formula (4) $$\bar s_k = s_k (1 + \Delta_k) \tag{4}$$ formula (6) $$\bar i = i + \eta_i^{(k)} \tag{6}$$

but there is sort of different in the code, the $\Delta_k$ and $\eta$ are multiplied by lambda_local and lambda_d respectively which (lambda_local and lambda_d) are assigned [0, 0.25, 0.5, 0.75, 1] in the SSRNET_train.py.

Why are they(lambda_local and lambda_d) exist? Thank you.

[shamangary]

Hello. lambda_local and lambda_d are basically hyperparameters for tuning dynamic indexing and dynamic scaling. I just want to make it more flexible. The final choice for both values is one.

[Pirate-Pianist]

Thank you.