Trouble understanding the theory

[mt40]

Hi there, I have read your paper and since I'm new to the field, I have trouble understanding the part in Section 3.5 about SIM(3) matching

we incorporate a depth residual rd which penalizes deviations in inverse depth between keyframes, allowing to directly estimate the scaled transformation between them

So how is depth residual related to the scale? How can it allow estimating the scaled transformation?

[srikar8]

@kantlove hi have you got the answer ?