1 unstable release
0.1.0 | Apr 30, 2023 |
---|
#10 in #yolo
14KB
161 lines
Getting started
This library relies on the tch crate for bindings to the C++ Libtorch API. A more detailed installation of this library can be found at the link to this repository.
Inference
use rusty_yolo;
use tch;
fn main() {
let device = tch::Device::cuda_if_available();
let yolo_model = rusty_yolo::YOLO::new("model.pt", 384, 640, device);
let mut original_image = tch::vision::image::load("/images/zidane.jpg").unwrap();
let results = yolo_model.predict(&original_image, 0.25, 0.35);
yolo_model.draw_rectangle(&mut original_image, &results);
tch::vision::image::save(&original_image, "images/result.jpg").unwrap();
}
Dependencies
~12MB
~242K SLoC