InspireFace4J

InspireFace4J provides a Java-Native binding using FFM to the InspireFace face detection library.

Supported features:

• Face detection (Boundingbox + Confidence)
• Face attribute extraction
• Face embedding extraction
• Face landmark extraction

Video processing can be by using the libary in combination with Video4j.

Limitations

Currently only AMD64 Linux is supported. Support for other platforms is not planned. CUDA support using TensorRT is currently not working.

Usage

<dependency>
  <groupId>io.metaloom.inspireface4j</groupId>
  <artifactId>inspireface4j</artifactId>
  <version>0.0.1-SNAPSHOT</version>
</dependency>

License

The code of this project is "Apache License" but the license of the models may be different. Please

The licensing of the open-source models employed by InspireFace adheres to the same requirements as InsightFace, specifying their use solely for academic purposes and explicitly prohibiting commercial applications.

Models

mkdir packs && cd packs
wget https://github.com/HyperInspire/InspireFace/releases/download/v1.x/Pikachu
wget https://github.com/HyperInspire/InspireFace/releases/download/v1.x/Megatron

# Not supported (Missing TensorRT support)
# wget https://github.com/HyperInspire/InspireFace/releases/download/v1.x/Megatron_TRT

Examples

Image Example

String imagePath = "src/test/resources/pexels-olly-3812743_1k_lower.jpg";

// Initialize video4j and InspirefaceLib (Video4j is used to handle OpenCV Mat)
Video4j.init();

try (InspirefaceSession session = InspirefaceLib.session(DEFAULT_PACK, 640,  ENABLE_FACE_RECOGNITION, ENABLE_FACE_ATTRIBUTE)) {

	// Load the image and invoke the detection
	BufferedImage img = ImageUtils.load(new File(imagePath));
	Mat imageMat = MatProvider.mat(img, Imgproc.COLOR_BGRA2BGR565);
	CVUtils.bufferedImageToMat(img, imageMat);

	// Invoke the detection
	FaceDetections detections = session.detect(imageMat, true);

	// Print the detections
	for (Detection detection : detections) {
		System.out.println(detection.box() + " @ " + detection.conf());
	}
	// Show the result and release the mat and the detection data
	ImageUtils.show(imageMat);
	MatProvider.released(imageMat);
}

Video Example

// Initialize video4j and InspirefaceLib (Video4j is used to handle OpenCV Mat)
Video4j.init();
SimpleImageViewer viewer = new SimpleImageViewer();


try (InspirefaceSession session = InspirefaceLib.session("packs/Pikachu", 640, ENABLE_FACE_RECOGNITION, ENABLE_FACE_ATTRIBUTE)) {

	// Open the video using Video4j
	try (VideoFile video = VideoFile.open("src/test/resources/8090198-hd_1366_720_25fps.mp4")) {

		// Process each frame
		VideoFrame frame;
		while ((frame = video.frame()) != null) {
			// System.out.println(frame);

			// Optionally downscale the frame
			CVUtils.resize(frame, 512);

			// Run the detection on the mat reference
			FaceDetections detections = session.detect(frame.mat(), true);

			if (!detections.isEmpty()) {
				// Extract the face embedding from the first face
				float[] embedding = session.embedding(frame.mat(), detections, 0);
				// Extract the face attributes
				List<FaceAttributes> attrs = session.attributes(frame.mat(), detections, true);
			}

			// Print the detections
			for (Detection detection : detections) {
				double confidence = detection.conf();
				BoundingBox box = detection.box();
				System.out.println("Frame[" + video.currentFrame() + "] = " + confidence + " @ " + box);
			}

			viewer.show(frame.mat());
		}
	}
}

Build

Requirements:

• InspireFace 1.2.1
• JDK 23 or newer
• Maven
• GCC 13

Building native code

# Download and extract inspireface-linux-x86-ubuntu18-1.2.1.zip from https://github.com/HyperInspire/InspireFace/releases
cd  inspireface4j
wget https://github.com/HyperInspire/InspireFace/releases/download/v1.2.1/inspireface-linux-x86-ubuntu18-1.2.1.zip
unp inspireface-linux-x86-ubuntu18-1.2.1.zip

cd jinspirelib
./build.sh

Notes for building from source

The CMakeLists.txt needs to be adapted to include all the different sources (e.g inspireface + inspirecv)

# Clone inspireface - my Head Rev: efb5639ec66d4e94004e4d16f34f44630179f95a
git clone [email protected]:HyperInspire/InspireFace.git
git clone [email protected]:deepinsight/insightface.git 

# Fix for minor path issue:
# cd insightface/cpp-package/inspireface/cpp/inspireface
# mv Initialization_module/ initialization_module/

CUDA support

I tried to build the project using TensorRT + CUDA but failed. I also had issues with inspireface-linux-tensorrt-cuda12.2_ubuntu22.04-1.2.1.zip.

Releasing

# Set release version and commit changes
mvn versions:set -DgenerateBackupPoms=false
git add pom.xml ; git commit -m "Prepare release"

# Invoke release
mvn clean deploy -Drelease