How can I get the output of yolov5 tflite model through "get_output<T: TensorType>(&self, name: &str)"

[alps-sun]

Summary

I want to get the output of yolov5 tflite model

#[wasmedge_bindgen]
pub fn count() {
    let start = Instant::now();

    let model_data: &[u8] = include_bytes!("yolov5s_counter.tflite");
    let labels = include_str!("classes.txt");

    // let image_path = "image.jpg";
    // let image_data = match std::fs::read(image_path) {
    //     Ok(data) => data,
    //     Err(error) => return Err(format!("Failed to read image file: {}", error)),
    // };

    // let flat_img = wasmedge_tensorflow_interface::load_jpg_image_to_rgb8(&image_data[..], 192, 192);
    let mut file_img = File::open("image.jpg").unwrap();
    let mut img_buf = Vec::new();
    file_img.read_to_end(&mut img_buf).unwrap();
    let flat_img = wasmedge_tensorflow_interface::load_jpg_image_to_rgb8(&img_buf, 192, 192);


    // The flat_img is a vec<f32> which contains normalized image in rgb32f format and resized to 224x224.
    println!("RUST: Loaded image in ... {:?}", start.elapsed());

    let mut session = wasmedge_tensorflow_interface::Session::new(
        &model_data,
        wasmedge_tensorflow_interface::ModelType::TensorFlowLite,
    );
    session
        .add_input("input", &flat_img, &[1, 192, 192, 3])
        .run();


    let output_tensor = session.get_output("output");
    println!("output_tensor: {:?}", output_tensor);
    println!("output-tensor   len: {}", output_tensor.len());

    // let label_probabilities: Vec<f32> = output_tensor.get_data::<f32>()?;
    let label_probabilities: Vec<f32> = output_tensor.to_vec();
    let labels_str = include_str!("classes.txt");
    let labels: Vec<&str> = labels_str.lines().collect();



    let person_count1 = label_probabilities[0];
    println!("Person count1: {}", person_count1);

    let mut person_count = 0;
    for i in 0..label_probabilities.len() {
        if label_probabilities[i] > 0.5 && labels[i] == "person" {
            person_count += 1;
        }
    }

    println!("Person count: {}", person_count);

    let ret_str: String;
    ret_str = format!("Person count: {}", person_count);

    println!(
        "RUST: Finished post-processing in ... {:?}",
        start.elapsed()
    );
    // return Ok(ret_str.as_bytes().to_vec());
}

Additional Details

 let output_tensor = session.get_output("output");

thank you