handtrack | Machine Vision Segmentation tool for extracting a mask | Computer Vision library

I am trying to upload my application on Oculus store but I get the following errors:

I have read the issues but none of them are included in my Manifest. I do not need these permissions either in my application. So how do I rectify something that is not already present in the manifest?

Being new to MediaPipe, I am not familiar with concept of graph, node, subgraph etc.

After building an aar file of BoxTracking, unable to run it within a Android Studio gradle based project due to some unknown input and output parameters required by model

On comparing HandTracking graph and BoxTracking graph using the visualizer tool and with a working project of HandTracking with aar file added as lib, I added new required input streams and side packets as seen in the graph.

Results are always some errors, mainly due to something wrong in inputs or BoxTracking being a subgraph which is used directly. How to know which input is required and data type of input to run this?

I am beginner in deep learning. I am using this dataset and I want my network to detect keypoints of a hand.

How can I make my output layer's nodes to be in range [-1, 1] (range of normalized 2D points)? Another problem is when I train for more than 1 epoch the loss gets negative values

criterion: torch.nn.MultiLabelSoftMarginLoss() and optimizer: torch.optim.SGD()

Here u can find my repo

I have tried to get hand mesh data from Hololens2 using MRTK V2 and Unity C#. Now, I can get hand mesh data with turning on Hand Mesh Visualization option and referring MRTK HandTracking guide.

Unfortunately, the visualization(drawing hand CG) is heavy workload. So, I would like to get hand mesh without turning on Hand Mesh Visualization option but OnHandMeshUpdated function is not called due to turning off Hand Mesh Visualization option.

Does anyone know how to get hand mesh data from Hololens2 without turning on Hand Mesh Visualization option?

Hand Mesh Visualization option

It is unclear to me how to properly stop and deallocate the MPPGraph. I created a framework very similar to this. Every time a dealloc is called in some way this exception is thrown. Thread 1: Exception: "waitUntilDoneWithError: should not be called on the main thread".

I don't know how to not call it in the main thread and was hoping someone had some insight on this.

Here you can find the swift code that is calling the mediapipe framework. This example has been created with the framework that can be found here.

So it would appear that this would be a simple solution, but as I can not find any documentation on how to do exactly this it's effectively the same as brute-force guessing a password.

• Unity Versions: 2019.3.1, 2019.3.4(current)
• Platform: Universal Windows Platform
• MRTK: 2.2, 2.3(current)
• HoloLens 2 OS: Windows Holographic Operating System
  
  

• I push a button and the file browser/explorer appears inside my Unity scene
  
  

• I can not launch the file browser/explorer in HoloLens 2.
  
  

With MRTK 2/HoloLens 2 you are able to launch external apps without exiting the Unity application. Something that HoloLens 1 could not do. Microsoft provides proof of this in their Unity examples package: Assets/MixedRealityToolkit.Examples/Demos/HandTracking/Scenes/HandInteraction.Examples.unity once you have loaded the .Foundation and .Examples external packages into your Unity project.

In the provided scene, out of all the presented object there are two buttons off to the right side that when pressed will launch the Edge Browser or the OS's settings application. This is accomplished with a launch URI attached script that runs .OpenURL on a string provided by the user via the GameObject's inspector.

And the code snippet (provided by Microsoft in MRTK2) that runs the user-inputted string: