usls 0.1.11

A Rust library integrated with ONNXRuntime, providing a collection of ML models.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141

use aksr::Builder;
use anyhow::Result;
use ndarray::Axis;
use rayon::prelude::*;

use crate::{elapsed_module, Config, DynConf, Engine, Hbb, Image, Keypoint, Processor, Xs, Y};

#[derive(Builder, Debug)]
pub struct RTMO {
    engine: Engine,
    height: usize,
    width: usize,
    batch: usize,

    spec: String,
    processor: Processor,
    confs: DynConf,
    kconfs: DynConf,
    names: Vec<String>,
}

impl RTMO {
    pub fn new(config: Config) -> Result<Self> {
        let engine = Engine::try_from_config(&config.model)?;
        let spec = engine.spec().to_string();
        let (batch, height, width) = (
            engine.batch().opt(),
            engine.try_height().unwrap_or(&512.into()).opt(),
            engine.try_width().unwrap_or(&512.into()).opt(),
        );
        let names: Vec<String> = config.keypoint_names().to_vec();
        let nk = config.nk().unwrap_or(17);
        let confs = DynConf::new_or_default(config.class_confs(), 1);
        let kconfs = DynConf::new_or_default(config.keypoint_confs(), nk);
        let processor = Processor::try_from_config(&config.processor)?
            .with_image_width(width as _)
            .with_image_height(height as _);

        Ok(Self {
            engine,
            height,
            width,
            batch,

            spec,
            processor,
            confs,
            kconfs,
            names,
        })
    }

    fn preprocess(&mut self, xs: &[Image]) -> Result<Xs> {
        Ok(self.processor.process_images(xs)?.into())
    }

    fn inference(&mut self, xs: Xs) -> Result<Xs> {
        self.engine.run(xs)
    }

    pub fn forward(&mut self, xs: &[Image]) -> Result<Vec<Y>> {
        let ys = elapsed_module!("RTMO", "preprocess", self.preprocess(xs)?);
        let ys = elapsed_module!("RTMO", "inference", self.inference(ys)?);
        let ys = elapsed_module!("RTMO", "postprocess", self.postprocess(ys)?);

        Ok(ys)
    }

    fn postprocess(&mut self, xs: Xs) -> Result<Vec<Y>> {
        let ys: Vec<Y> = xs[0]
            .axis_iter(Axis(0))
            .into_par_iter()
            .zip(xs[1].axis_iter(Axis(0)).into_par_iter())
            .enumerate()
            .map(|(idx, (batch_bboxes, batch_kpts))| {
                let (height_original, width_original) = (
                    self.processor.images_transform_info[idx].height_src,
                    self.processor.images_transform_info[idx].width_src,
                );
                let ratio = self.processor.images_transform_info[idx].height_scale;

                let mut y_bboxes = Vec::new();
                let mut y_kpts: Vec<Vec<Keypoint>> = Vec::new();
                for (xyxyc, kpts) in batch_bboxes
                    .axis_iter(Axis(0))
                    .zip(batch_kpts.axis_iter(Axis(0)))
                {
                    // bbox
                    let x1 = xyxyc[0] / ratio;
                    let y1 = xyxyc[1] / ratio;
                    let x2 = xyxyc[2] / ratio;
                    let y2 = xyxyc[3] / ratio;
                    let confidence = xyxyc[4];

                    if confidence < self.confs[0] {
                        continue;
                    }
                    y_bboxes.push(
                        Hbb::default()
                            .with_xyxy(
                                x1.max(0.0f32).min(width_original as _),
                                y1.max(0.0f32).min(height_original as _),
                                x2,
                                y2,
                            )
                            .with_confidence(confidence)
                            .with_id(0)
                            .with_name("Person"),
                    );

                    // keypoints
                    let mut kpts_ = Vec::new();
                    for (i, kpt) in kpts.axis_iter(Axis(0)).enumerate() {
                        let x = kpt[0] / ratio;
                        let y = kpt[1] / ratio;
                        let c = kpt[2];
                        if c < self.kconfs[i] {
                            kpts_.push(Keypoint::default());
                        } else {
                            let mut kpt_ =
                                Keypoint::default().with_id(i).with_confidence(c).with_xy(
                                    x.max(0.0f32).min(width_original as _),
                                    y.max(0.0f32).min(height_original as _),
                                );

                            if !self.names.is_empty() {
                                kpt_ = kpt_.with_name(&self.names[i]);
                            }

                            kpts_.push(kpt_);
                        }
                    }
                    y_kpts.push(kpts_);
                }
                Y::default().with_hbbs(&y_bboxes).with_keypointss(&y_kpts)
            })
            .collect();

        Ok(ys)
    }
}