uoa-eresearch / hololens_facial_recognition

A Unity hololens app to detect faces and display their attributes

hololens_facial_recognition

Facial recognition for the Microsoft Hololens.

Click the Main Camera in Unity and edit Face Api Url if you need to point it to a different running instance of https://github.com/UoA-eResearch/face_api
Air tap or click with the clicker to take a photo/post it to the Face API.

hololens_facial_recognition's Issues

Metadata file Assembly-CSharp-firstpass.dll not found

In visual studio, when I lanch the app in Release mode, x86, Device:

• Metadata file Assembly-CSharp-firstpass.dll not found
• Could not write lines to "obj\x86\Release\Release\Assembly-CSharp-firstpass.csprojResolveAssemblzReference.cache". the specified path, filename or both are too long. The fully qualified file name must be less than 260 characters, and the directory name must be less than 248 characters.

FR: enhance & move output

After making it work, I think others may stumble upon the same points as me, that 1st) there is no meaningful output when the REST call doesn't work for some reason. And 2nd) it would be great, if the output follows my view instead of staying at a fix position.
The 1st is propably simple, but I have yet no idea if the 2nd is.
And an additional basic question: Do I always have to go from the Unity project, or can I stick to the Visual C# solution once it's exported for the 1st time? (by the way, it may make sense to add to the readme, what you told me to use as base tutorial for it).

Softwarebitmap post to web api

Hi,
I am currently working on something similar to this project, where I have live feed from hololens camera and I use windows sdk to detect faces and draw rectangles around the faces in the 3d view. whenever I discover a face, I need to send its frame to an API to do face recognition the same way you do. what am facing is the byre array that am getting from the capture element is a Softwarebitmap, and my API is an ASP.NET web API. I receive the post request and I get the binary data, but I can't manage to convert it to a Bitmap to be used in my face recognition library. do you have any idea how to do so.?

App doesn't handle being minimised

Minimising the app via the bloom gesture and reopening it sometimes results in a stuck state of trying to map the room

problem in building the app

I have this problem

after I:
I've replaced the API keys just like you've suggested
Git cloned your project and load it into Unity
Opened the 'Main' scene
Changed the build settings: SDK--> windows 10, UWP build type --> D3D, Debugging -- >Unity c# project
Build it
Open up the sln file
debug-> start debugging

Prompt the user to air tap

https://github.com/sizingservers/hololens_facial_recognition/commit/b986c66a8ea9ad0ae697f32e611d030044bfd412 adds a transparent icon of a camera, and an animation prompting the user to air tap. This would solve the issue of the user being confused by not seeing anything when first putting on the hololens. It would be good to include this feature in this repository, though without the branding

While the app is running in Hololens, cannot perform live preview on browser at the same time.

Hi,
Your project works great and modified it to include face identification as well using face API.
The thing is I want to demo it to a group of people but the app crashes when I start the live preview in the browser. When i looked up for the solution, i found that this is the limitation of MRC https://developer.microsoft.com/en-us/windows/mixed-reality/mixed_reality_capture_for_developers#photo.2Fvideo_camera_access
Would like to know if there is any work around for this problem?

project.lock.json workaround

Half a year ago I had problems with compiling the code. Here follows a workaround that might be useful for someone:

• Rebuild the solution
• If you get an error on project.lock.json, find it in the App\HoloWorld folder and open it using a text editor
• Find and replace all occurances of UAP,Version=v10.0.10240 with UAP,Version=v10.0
• Build, but not rebuild, the solution
• Party

No results returned?

When I run the solution and airtap, I see red text "Photo captured, processing..."
But nothing is shown after that. I've copied config.cfg.example, renamed it to config.cfg, and updated the API keys within (as well as within GazeGestureManager.cs itself).

I have also ensured that Microphone is ticked in Capabilities.

Where are Azure API Endpoints defined? Could that be my issue?

EDIT: Found the problem. Indeed it was the Endpoint URLs. In GazeGestureManager.cs, I had to change the two "api.projectoxford.ai" URLs to the Endpoints that are shown on my Azure APIs page.

Built the app but nothing happened

Hi,

I was attempting to build the facial recognition app but I couldn't get the app to work. Just to be sure if the build steps are correct, this is what I did:

1. I've replaced the API keys just like you've suggested
2. Git cloned your project and load it into Unity
3. Opened the 'Main' scene
4. Changed the build settings: SDK--> windows 10, UWP build type --> D3D, Debugging -- >Unity c# project
5. Build it
6. Open up the sln file
7. Edited the package.appxmanifest, basically changing the TargetDeviceFamily and MaxVersionTested entries
8. Changed the deployment target entries
9. Build the project, deployed,
10. Hololens shows the unity splash screen, point my hololens toward a picture of someone's face on my monitor screen but nothing happened.

My question is what went wrong?

-Ken

我想知道在进行点击手势后多久人脸信息可以显示出来？

是几乎实时、一秒、两秒、三秒或更长，粗略估计一下就好

Recognition jobs unhandled error

Recognition jobs response for loop should handle the case where the response is an error - otherwise jobs further in the queue which were a success won't show. Continue statement should suffice

Unable to get the project to work

Hi neon_ninja,

I am coming across multiple issues trying to run this. Not sure what I am doing wrong.

When I am running using HoloLens version of Unity 5.4.0.f3. When I run it does say loaded settings from config.cfg where in I have populated my API keys. However, when I run in Unity all it does is gives me a cropped image of my example.jpg. Also on my Unity console it gives me

Also when I load the solution into VS2015, I am unable to explore the Configuration properties where I could change my target device. Any help on this would be greatly appreciated.

Some errors

I built it in Unity 5.5.0f3 , some errors happened.

1. in GazeGestureManager.cs line 45.
   captureObject.StartPhotoModeAsync(c, false ,OnPhotoModeStarted);
   it should be `captureObject.StartPhotoModeAsync(c,OnPhotoModeStarted);
   2.I think you should add some guide in Readme.md,such as user should put config.cfg under Assets folder.

anyway,this project helps me a lot, thank you very much.(sorry,my english is so bad).

如果能有一个对框架的简单介绍会更好

我还没有实现你的工程，我想知道他是怎么运行的。