aprender-core 0.29.1

Next-generation machine learning library in pure Rust
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
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238


    #[test]
    fn test_spectral_subtraction_debug_clone() {
        let isolator = SpectralSubtractionIsolator::default();
        let cloned = isolator.clone();
        let debug_str = format!("{isolator:?}");
        assert!(!debug_str.is_empty());
        assert_eq!(cloned.over_subtraction(), isolator.over_subtraction());
    }

    #[test]
    fn test_spectral_subtraction_isolate_short_audio() {
        // Audio shorter than noise_frames * fft_size
        let isolator = SpectralSubtractionIsolator::default();
        let short_audio: Vec<f32> = (0..100).map(|i| (i as f32 * 0.01).sin()).collect();
        let result = isolator.isolate(&short_audio);
        assert!(result.is_ok());
    }

    #[test]
    fn test_spectral_subtraction_isolate_with_profile_empty() {
        let isolator = SpectralSubtractionIsolator::default();
        let profile = NoiseProfile::from_frames(&[vec![0.1; 257]], 16000);
        let result = isolator.isolate_with_profile(&[], &profile);
        assert!(result.is_err());
        match result.unwrap_err() {
            VoiceError::InvalidAudio(msg) => assert!(msg.contains("Empty")),
            other => panic!("Expected InvalidAudio, got {other:?}"),
        }
    }

    #[test]
    fn test_spectral_subtraction_estimate_noise_empty() {
        let isolator = SpectralSubtractionIsolator::default();
        let result = isolator.estimate_noise(&[]);
        assert!(result.is_err());
        match result.unwrap_err() {
            VoiceError::InvalidAudio(msg) => assert!(msg.contains("Empty")),
            other => panic!("Expected InvalidAudio, got {other:?}"),
        }
    }

    #[test]
    fn test_wiener_filter_with_smoothing() {
        let isolator = WienerFilterIsolator::default();
        // Normal value
        let filtered = isolator.with_smoothing(0.5);
        assert!((filtered.smoothing - 0.5).abs() < f32::EPSILON);

        // Clamp low
        let low = WienerFilterIsolator::default().with_smoothing(-1.0);
        assert!((low.smoothing - 0.0).abs() < f32::EPSILON);

        // Clamp high
        let high = WienerFilterIsolator::default().with_smoothing(5.0);
        assert!((high.smoothing - 1.0).abs() < f32::EPSILON);
    }

    #[test]
    fn test_wiener_compute_gain_zero_noise() {
        let gain = WienerFilterIsolator::compute_gain(1.0, 0.0);
        assert!((gain - 1.0).abs() < f32::EPSILON);

        let gain_neg = WienerFilterIsolator::compute_gain(1.0, -1.0);
        assert!((gain_neg - 1.0).abs() < f32::EPSILON);
    }

    #[test]
    fn test_wiener_compute_gain_equal_power() {
        let gain = WienerFilterIsolator::compute_gain(1.0, 1.0);
        assert!((gain - 0.5).abs() < 1e-6);
    }

    #[test]
    fn test_wiener_filter_isolate_empty() {
        let isolator = WienerFilterIsolator::default();
        let result = isolator.isolate(&[]);
        assert!(result.is_err());
    }

    #[test]
    fn test_wiener_filter_isolate_with_profile_empty() {
        let isolator = WienerFilterIsolator::default();
        let profile = NoiseProfile::from_frames(&[vec![0.1; 257]], 16000);
        let result = isolator.isolate_with_profile(&[], &profile);
        assert!(result.is_err());
    }

    #[test]
    fn test_wiener_filter_estimate_noise_empty() {
        let isolator = WienerFilterIsolator::default();
        let result = isolator.estimate_noise(&[]);
        assert!(result.is_err());
    }

    #[test]
    fn test_wiener_filter_isolate_short_audio() {
        // Audio shorter than noise_frames * fft_size
        let isolator = WienerFilterIsolator::default();
        let short_audio: Vec<f32> = (0..100).map(|i| (i as f32 * 0.01).sin()).collect();
        let result = isolator.isolate(&short_audio);
        assert!(result.is_ok());
        let isolated = result.expect("isolation failed");
        assert!(!isolated.audio.is_empty());
    }

    #[test]
    fn test_wiener_filter_debug_clone() {
        let isolator = WienerFilterIsolator::default();
        let cloned = isolator.clone();
        let debug_str = format!("{isolator:?}");
        assert!(!debug_str.is_empty());
        assert_eq!(cloned.smoothing, isolator.smoothing);
    }

    #[test]
    fn test_estimate_snr_short_audio() {
        // Audio shorter than frame_size (256)
        let audio: Vec<f32> = (0..100).map(|i| (i as f32 * 0.01).sin()).collect();
        let snr = estimate_snr(&audio);
        // num_frames = 100/100 = 1 (since frame_size is min(256, 100) = 100)
        assert!(snr.is_finite());
    }

    #[test]
    fn test_estimate_snr_zeros() {
        let audio = vec![0.0_f32; 1000];
        let snr = estimate_snr(&audio);
        // With all zeros, signal_rms and noise_rms are both 0
        assert_eq!(snr, 0.0);
    }

    #[test]
    fn test_estimate_snr_single_sample() {
        let audio = vec![1.0_f32];
        let snr = estimate_snr(&audio);
        // frame_size = min(256, 1) = 1, num_frames = 1/1 = 1
        assert!(snr.is_finite());
    }

    #[test]
    fn test_spectral_entropy_all_zeros() {
        let zeros = vec![0.0_f32; 100];
        let entropy = spectral_entropy(&zeros);
        // sum <= 0.0 => return 0.0
        assert_eq!(entropy, 0.0);
    }

    #[test]
    fn test_spectral_entropy_single_element() {
        let single = vec![1.0_f32];
        let entropy = spectral_entropy(&single);
        // max_entropy = ln(1) = 0.0, returns 0.0
        assert_eq!(entropy, 0.0);
    }

    #[test]
    fn test_spectral_entropy_negative_values() {
        // Negative values sum to zero
        let negs = vec![-1.0_f32; 10];
        let entropy = spectral_entropy(&negs);
        // sum = -10.0 which is <= 0.0
        assert_eq!(entropy, 0.0);
    }

    #[test]
    fn test_detect_voice_activity_zero_frame_size() {
        let audio = vec![1.0_f32; 100];
        let vad = detect_voice_activity(&audio, 0, 0.1);
        assert!(vad.is_empty());
    }

    #[test]
    fn test_detect_voice_activity_all_speech() {
        let loud_audio = vec![1.0_f32; 1000];
        let vad = detect_voice_activity(&loud_audio, 100, 0.01);
        assert!(vad.iter().all(|&v| v));
    }

    #[test]
    fn test_detect_voice_activity_all_silence() {
        let quiet_audio = vec![0.0_f32; 1000];
        let vad = detect_voice_activity(&quiet_audio, 100, 0.1);
        assert!(vad.iter().all(|&v| !v));
    }

    #[test]
    fn test_detect_voice_activity_partial_frame() {
        // Audio length not evenly divisible by frame_size
        let audio = vec![1.0_f32; 150];
        let vad = detect_voice_activity(&audio, 100, 0.01);
        // (150 + 100 - 1) / 100 = 2 frames
        assert_eq!(vad.len(), 2);
    }

    #[test]
    fn test_spectral_subtraction_config_accessor() {
        let config = IsolationConfig::aggressive();
        let isolator = SpectralSubtractionIsolator::new(config.clone());
        assert_eq!(
            isolator.config().reduction_strength,
            config.reduction_strength
        );
    }

    #[test]
    fn test_wiener_filter_config_accessor() {
        let config = IsolationConfig::mild();
        let isolator = WienerFilterIsolator::new(config.clone());
        assert_eq!(
            isolator.config().reduction_strength,
            config.reduction_strength
        );
    }

    #[test]
    fn test_wiener_filter_estimate_noise_valid() {
        let isolator = WienerFilterIsolator::default();
        let noise: Vec<f32> = (0..4000).map(|i| 0.01 * (i as f32 * 0.005).sin()).collect();
        let profile = isolator.estimate_noise(&noise);
        assert!(profile.is_ok());
        let p = profile.expect("estimation failed");
        assert!(p.is_valid());
    }

    #[test]
    fn test_wiener_filter_isolate_with_profile_valid() {
        let isolator = WienerFilterIsolator::default();
        let audio: Vec<f32> = (0..8000).map(|i| (i as f32 * 0.001).sin()).collect();
        let noise: Vec<f32> = (0..4000).map(|i| 0.01 * (i as f32 * 0.005).sin()).collect();
        let profile = isolator
            .estimate_noise(&noise)
            .expect("noise estimation failed");
        let result = isolator.isolate_with_profile(&audio, &profile);
        assert!(result.is_ok());
        let isolated = result.expect("isolation failed");
        assert_eq!(isolated.audio.len(), audio.len());
    }