polyvoice 0.1.0

Speaker diarization library for Rust — online and offline, ONNX-powered, ecosystem-agnostic
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

//! Math utilities for diarization.

/// Compute cosine similarity between two vectors.
///
/// Returns a value in `[-1.0, 1.0]`. Returns `0.0` if either vector is near-zero.
pub fn cosine_similarity(a: &[f32], b: &[f32]) -> f32 {
    assert_eq!(a.len(), b.len(), "vector length mismatch");
    let mut dot = 0.0f32;
    let mut norm_a = 0.0f32;
    let mut norm_b = 0.0f32;
    for (&x, &y) in a.iter().zip(b.iter()) {
        dot += x * y;
        norm_a += x * x;
        norm_b += y * y;
    }
    if norm_a < 1e-8 || norm_b < 1e-8 {
        return 0.0;
    }
    dot / (norm_a.sqrt() * norm_b.sqrt())
}

/// L2-normalize a vector in-place.
pub fn l2_normalize(vec: &mut [f32]) {
    let norm: f32 = vec.iter().map(|x| x * x).sum::<f32>().sqrt();
    if norm > 1e-8 {
        for v in vec.iter_mut() {
            *v /= norm;
        }
    }
}

/// Compute the mean vector of a set of vectors.
pub fn mean_vector(vectors: &[Vec<f32>]) -> Option<Vec<f32>> {
    if vectors.is_empty() {
        return None;
    }
    let dim = vectors[0].len();
    let mut sum = vec![0.0f32; dim];
    for v in vectors {
        for (s, &x) in sum.iter_mut().zip(v.iter()) {
            *s += x;
        }
    }
    let n = vectors.len() as f32;
    for s in &mut sum {
        *s /= n;
    }
    Some(sum)
}

/// Moving average over a slice.
pub fn moving_average(data: &[f32], window: usize) -> Vec<f32> {
    if window == 0 || data.is_empty() {
        return data.to_vec();
    }
    let mut result = Vec::with_capacity(data.len());
    let half = window / 2;
    for i in 0..data.len() {
        let start = i.saturating_sub(half);
        let end = (i + half + 1).min(data.len());
        let avg = data[start..end].iter().sum::<f32>() / (end - start) as f32;
        result.push(avg);
    }
    result
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_cosine_similarity_identical() {
        let a = vec![1.0, 2.0, 3.0];
        let b = vec![1.0, 2.0, 3.0];
        let sim = cosine_similarity(&a, &b);
        assert!((sim - 1.0).abs() < 1e-5);
    }

    #[test]
    fn test_cosine_similarity_orthogonal() {
        let a = vec![1.0, 0.0, 0.0];
        let b = vec![0.0, 1.0, 0.0];
        let sim = cosine_similarity(&a, &b);
        assert!(sim.abs() < 1e-5);
    }

    #[test]
    fn test_cosine_similarity_opposite() {
        let a = vec![1.0, 2.0, 3.0];
        let b = vec![-1.0, -2.0, -3.0];
        let sim = cosine_similarity(&a, &b);
        assert!((sim + 1.0).abs() < 1e-5);
    }

    #[test]
    fn test_l2_normalize() {
        let mut v = vec![3.0, 4.0];
        l2_normalize(&mut v);
        assert!((v[0] - 0.6).abs() < 1e-5);
        assert!((v[1] - 0.8).abs() < 1e-5);
    }

    #[test]
    fn test_mean_vector() {
        let vectors = vec![vec![1.0, 2.0], vec![3.0, 4.0]];
        let mean = mean_vector(&vectors).unwrap();
        assert!((mean[0] - 2.0).abs() < 1e-5);
        assert!((mean[1] - 3.0).abs() < 1e-5);
    }
}