Mismatch in inference on image used in paper

Hi, this is awesome work! Congratulations!

I tried to make an inference on one of the photos used in the paper figures ([attached](https://imgur.com/e8rUw1n)). The depth map I generated [(attached)](https://imgur.com/RN9M4vW) using the pre-trained model looks very different from the depth map published in the paper (mine almost doesn't have anu Blue color). Also, I used the function for generating point cloud to estimate the metric depth of the marked point in the image. Whereas the paper mentions a metric depth of 5.82 meters, I am getting 3.5 meters.

Could you please help me with what I may be doing wrong? Any pointers would be highly appreciated.

Thanks!