mpottinger

Excellent, thanks for this. Thought I may need to implement it myself.

I too am interested in training this. It is just what we need for custom object detection. Generic multi-class object detectors like Yolo are great for object detection competitions, but...

> Thank you very much for your prompt response. > > what do you mean by > > > you have to align each prediction using some known 3D points...

@zengbw @lhanaf Yes I am also interested of course! It said coming soon but so far nothing. Maybe commercially bought out? It looks like an amazing 3d scanning system. It...

@mpottinger Changing the loops to iterate until 3 instead of 4 worked for me. Depth is aligned nicely to RGB. Don't know why I had to make the change, but...

@mpottinger FYI caching works. I just initialize a very large static array with precalculated values and I get very nice real time framerates. It is a huge lookup table though,...

@mpottinger Just so anyone needs the same thing in the future, since I had a bit of a struggle, I have it solved for my use (real time low latency...

I will have to take a look at my code when I have time. The sizes in pixels of the depth and rgb are going to be different because the...

@promethe42 Im not the author but I just took a look and I did not see any uint8_t for the depth in the code. Strange. 16 bit depth is the...

@promethe42 ok yes now I see it. I don't know why either. The code also had other issues for me that I had to overcome.