Unity NullReferenceException in running NatML handpose package

Question

I am trying to run a hand pose estimator unitypackage from NatML. The original code was taking Texture 2D Image as input and results displayed were correct. However, I wanted to detect the hand in real time and display the results in real time too. In unity, there is no direct conversion from webcamtexture to texture2D so I extracted pixels from webcam and gave it as input to my model to run. I got no errors in unity with this following code but when I press play, it says NullReferenceException I have also attached image of my unity editor. Can anyone please guide me what possibly could be the error? namespace NatSuite.Examples {

    using UnityEngine;
    using NatSuite.ML;
    using NatSuite.ML.Features;
    using NatSuite.ML.Vision;                      
    using NatSuite.ML.Visualizers;
    using Stopwatch = System.Diagnostics.Stopwatch;
    using System.Collections;
    using System.Collections.Generic;

    public sealed class HandPoseSample : MonoBehaviour {
        
        [Header(@"NatML Hub")]
        public string accessKey;

        [Header(@"Prediction")]
        public WebCamTexture webcam;
        Color32[] data;
        public Texture2D output;
        public HandPoseVisualizer visualizer;
        //public HandPosePredictor predictor;
        public MLModelData modelData;
        async void Start() {
            webcam = new WebCamTexture();
            webcam.Play();
            output = new Texture2D(webcam.width, webcam.height);
            data = new Color32[webcam.width * webcam.height];
            // Fetch the model data from Hub
            Debug.Log("Fetching model data from Hub");
            modelData = await MLModelData.FromHub("@natsuite/hand-pose", accessKey);
            
            
            // Deserialize the model
            var model = modelData.Deserialize();
            // Create the hand pose predictor
            var predictor = new HandPosePredictor(model);
            // Create input feature
            var input = new MLImageFeature(output);
            // Predict
            var watch = Stopwatch.StartNew();
            var hand = predictor.Predict(input);
            watch.Stop();
            // Visualize
            Debug.Log($"Detected {hand.handedness} hand with score {hand.score:0.##} after {watch.Elapsed.TotalMilliseconds}ms");
            visualizer.Render(hand);
            // Dispose the model
            model.Dispose();
        }
        void Update () {
            if (data != null) {
                webcam.GetPixels32(data);
                output.SetPixels32(data);
                output.Apply();
            }
        }
    }
}

enter image description here