demonkenny/MAXINE-AR-SDK
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge pull request #2 from NVIDIA/jdsouza-dev

Browse Source 
Bug fixes, and updated README and documentation
This commit is contained in:
jdsouza90
2020-03-25 22:33:28 -07:00
committed by GitHub
parent 479970b1c0 a4fedf503b
commit 929d99bd1a
5 changed files with 13 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

BIN
NVIDIA AR SDK Programming Guide.pdf
View File

Binary file not shown.

4
samples/FaceTrack/FaceTrack.cpp
View File

@@ -755,6 +755,10 @@ DoApp::Err DoApp::initOfflineMode(const char *inputFilename, const char *outputF
DoApp::Err DoApp::fitFaceModel() {
DoApp::Err doErr = errNone;
nvErr = face_ar_engine.fitFaceModel(frame);
if (FLAG_captureOutputs) {
writeFrameAndEstResults(frame, face_ar_engine.output_bboxes, face_ar_engine.getLandmarks());
writeVideoAndEstResults(frame, face_ar_engine.output_bboxes, face_ar_engine.getLandmarks());
}
if (FaceEngine::Err::errNone == nvErr) {
#ifdef VISUALIZE

BIN
samples/FaceTrack/FaceTrack.exe
View File

Binary file not shown.

16
samples/FaceTrack/Readme.txt
View File

@@ -1,5 +1,5 @@
FaceTrack is a sample Windows application that demonstrates the face detection, landmark detection,
and 3D face reconstruction features of the NVIDIA AR SDK. The application requires a video feed
FaceTrack is a sample Windows application that demonstrates the face tracking, landmark tracking,
and 3D mesh tracking features of the NVIDIA AR SDK. The application requires a video feed
from a camera connected to the computer running the application, or from a video file, as specified
with command-line arguments (enumerated by executing: FaceTrack.exe --help).
@@ -7,15 +7,17 @@ The sample application can run in 3 modes which can be toggled through the '1',
the keyboard
1 - Face tracking
2 - Facial landmark tracking
3 - 3D face reconstruction
3 - 3D mesh tracking
For more controls and configurations of the sample app, please read the SDK programming guide.
3D face fitting requires the model file face_model0.nvf which is not part of the SDK package.
It has to be generated through the steps described below. The sample app can still be used in
mode 1 and 2 without this file.
3D mesh tracking feature requires a 3D Morphable Face Model (3DMM).
NVIDIA AR SDK does not include a 3DMM. Therefore, if you are using the 3D mesh tracking feature, you must configure NVIDIA AR SDK with 3DMM.
Note: To configure NVIDIA AR SDK with 3DMM, you can use the Surrey Face Model, as described in the steps below, or your own model.
The sample app can still be used in mode 1 and 2 without this file.
To generate the face model file face_model0.nvf required for 3D face fitting:
NVIDIA AR SDK provides the ConvertSurreyFaceModel.exe utility to convert 3DMM files to the NVIDIA .nvf format that is required by the SDK.
To generate the face model file face_model0.nvf required for 3D mesh tracking:
1) Download the following Surrey Face Model files from the eos project page on Github:
sfm_shape_3448.bin

BIN
tools/ConvertSurreyFaceModel.exe
View File

Binary file not shown.