hietalajulius

class ViewController: UIViewController, AVCaptureVideoDataOutputSampleBufferDelegate {
    var bufferSize: CGSize = .zero
    var inferenceTime: CFTimeInterval  = 0;
    
    private let session = AVCaptureSession()
    
    @IBOutlet weak var previewView: UIView!
    var rootLayer: CALayer! = nil
    private var previewLayer: AVCaptureVideoPreviewLayer! = nil
    private var detectionLayer: CALayer! = nil

hietalajulius

override func viewDidLoad() {
        super.viewDidLoad()
        setupCapture()
        setupOutput()
        setupLayers()
        try? setupVision()
        session.startRunning()
}

hietalajulius

 func setupCapture() {
        var deviceInput: AVCaptureDeviceInput!
        let videoDevice = AVCaptureDevice.DiscoverySession(deviceTypes: [.builtInWideAngleCamera], mediaType: .video, position: .back).devices.first
        do {
            deviceInput = try AVCaptureDeviceInput(device: videoDevice!)
        } catch {
            print("Could not create video device input: \(error)")
            return
        }
        

hietalajulius

func setupLayers() {
        previewLayer = AVCaptureVideoPreviewLayer(session: session)
        previewLayer.videoGravity = AVLayerVideoGravity.resizeAspectFill
        rootLayer = previewView.layer
        previewLayer.frame = rootLayer.bounds
        rootLayer.addSublayer(previewLayer)
        
        inferenceTimeBounds = CGRect(x: rootLayer.frame.midX-75, y: rootLayer.frame.maxY-70, width: 150, height: 17)
        
        inferenceTimeLayer = createRectLayer(inferenceTimeBounds, [1,1,1,1])

hietalajulius

 func setupVision() throws {
        guard let modelURL = Bundle.main.url(forResource: "yolov5n", withExtension: "mlmodelc") else {
            throw NSError(domain: "ViewController", code: -1, userInfo: [NSLocalizedDescriptionKey: "Model file is missing"])
        }
        
        do {
            let visionModel = try VNCoreMLModel(for: MLModel(contentsOf: modelURL))
            let objectRecognition = VNCoreMLRequest(model: visionModel, completionHandler: { (request, error) in
                DispatchQueue.main.async(execute: {
                    if let results = request.results {

hietalajulius

 func captureOutput(_ output: AVCaptureOutput, didOutput sampleBuffer: CMSampleBuffer, from connection: AVCaptureConnection) {
        guard let pixelBuffer = CMSampleBufferGetImageBuffer(sampleBuffer) else {
            return
        }
        
        let imageRequestHandler = VNImageRequestHandler(cvPixelBuffer: pixelBuffer, orientation: .right, options: [:])
        do {
            let start = CACurrentMediaTime()
            try imageRequestHandler.perform(self.requests)
            inferenceTime = (CACurrentMediaTime() - start)

hietalajulius

use actix::{Actor, Context, Handler};
use actix_broker::{BrokerIssue, BrokerSubscribe};
use actix_rt::System;
use std::println;
#[derive(actix::Message, Clone, Debug)]
#[rtype(result = "()")]
struct Message1 {
    pub level: i32,
}
#[derive(actix::Message, Clone, Debug)]

hietalajulius

// Initialize the random number generator
let mut rng = StdRng::seed_from_u64(123);
// ...
// Set the batch size for training
let batch_size = 16;
// ...
// Generate random input samples within the range [-3.0, 3.0]
let x: Array2<f64> = Array2::from_shape_fn((batch_size, 2), |_| rng.gen_range(-3.0..=3.0));

// Compute the target values for the input samples using the PDF function

hietalajulius

// Create a new neural network with 2 input features, 8 hidden units in each of the two hidden layers, and 1 output feature
let mut nn = NN::new(2, 8, 8, 1, &mut rng);
// ...
// Perform a forward pass through the neural network to compute the predicted output
let y = nn.forward(&x);

hietalajulius

// Create a mean squared error (MSE) loss function
let mut loss = MSE::new();
// ...
// Compute the loss value using the MSE loss function
let loss_value = loss.forward(&y, &target);