[Question] PreProcessor methods missing , e.g. scale()

[hjander]

i am trying to create a Yolov10 example with openvino-rs 0.7.2

I am trying to replicate this cpp code from here.

void Inference::InitialModel(const std::string &model_path) {
	ov::Core core;
	std::shared_ptr<ov::Model> model = core.read_model(model_path);

	if (model->is_dynamic()) {
		model->reshape({1, 3, static_cast<long int>(model_input_shape_.height), static_cast<long int>(model_input_shape_.width)});
	}

	ov::preprocess::PrePostProcessor ppp = ov::preprocess::PrePostProcessor(model);

  ppp.input().tensor().set_element_type(ov::element::u8).set_layout("NHWC").set_color_format(ov::preprocess::ColorFormat::BGR);
  ppp.input().preprocess().convert_element_type(ov::element::f32)
     .convert_color(ov::preprocess::ColorFormat::RGB).scale({ 255, 255, 255 });
  ppp.input().model().set_layout("NCHW");
  ppp.output().tensor().set_element_type(ov::element::f32);

  model = ppp.build();
	compiled_model_ = core.compile_model(model, "AUTO");
	inference_request_ = compiled_model_.create_infer_request();

	short width, height;

  const std::vector<ov::Output<ov::Node>> inputs = model->inputs();
  const ov::Shape input_shape = inputs[0].get_shape();

	height = input_shape[1];
	width = input_shape[2];
	model_input_shape_ = cv::Size2f(width, height);

  const std::vector<ov::Output<ov::Node>> outputs = model->outputs();
  const ov::Shape output_shape = outputs[0].get_shape();

	height = output_shape[1];
	width = output_shape[2];
	model_output_shape_ = cv::Size(width, height);
}

It seems the convert_color() and scale() methods are missing in openvino-rs PreProcessor though they exist in openvino-sys.

Are there any plans to add those methods or is there a particular method reason not to add them ?

[abrown]

I am completely open to adding those methods if you want to take a stab at it. In these 0.7.* releases we were primarily interested in switching over to OpenVINO's new 2.0 API and not necessarily completeness.

[j99ca]

@abrown I am also running into this issue. Is there any plan to extend the preprocessing functionality to include scale() and mean()?

[abrown]

Sure, contributions to add more methods to cover missing parts of the API are always welcome! Are you interested?