EfficientWord-Net-Trainer/simple_model.py at master · Ant-Brain/EfficientWord-Net-Trainer · GitHub

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
import torchvision
import torch
import torch.nn as nn
import torch.nn.functional as F
from resnet_custom import attentive_resnet50
import math

class EmptyModule(nn.Module):
    def __init__(self):
        super(EmptyModule, self).__init__()
    def forward(self,x):
        return x

class ResNet50_Classifier(nn.Module):
    def __init__(self, class_count:int) :
        super(ResNet50_Classifier, self).__init__()

        self.inp_batchnorm = nn.BatchNorm2d(1)

        self.feature_network = torchvision.models.resnet50()
        self.feature_network.conv1 = nn.Conv2d(1, 64, kernel_size=7, stride=2, padding=3, bias=False)
        self.feature_network.fc = nn.Linear(2048, class_count)

    def forward(self,x) :
        tmp = self.inp_batchnorm(x)
        #print(self.feature_network.bn1(tmp).shape)
        return self.feature_network(tmp)

class ResNetArc_Classifier(nn.Module):
    def __init__(self, model_type:str="resnet50", class_count:int=1000):
        super(ResNetArc_Classifier, self).__init__()
        resnet_model_fn = getattr(torchvision.models, model_type)
        self.inp_batch_norm = nn.BatchNorm2d(1)
        self.feature_network = resnet_model_fn()
        self.feature_network.conv1 = nn.Conv2d(1, 64, kernel_size=7, stride=2, padding=3, bias=False)
        self.feature_network.fc = nn.Identity()

        self.bn_final = nn.BatchNorm1d(2048)

        self.mapper = nn.Linear(2048, class_count, bias=False)

        self.s = 8
        self.m = 0
        self.update_margin(self.m, self.s)


    def update_margin(self, m = 0 , s=8):
        self.s = s
        self.m = m
        self.cos_m = math.cos(m)
        self.sin_m = math.sin(m)

    def forward(self, x ,y=None):
        assert self.s!=0

        x_norm = self.inp_batch_norm(x)
        feat = self.feature_network(x_norm)
        feat_norm = self.bn_final(feat)
        feat_l2 = F.normalize(feat_norm, dim=1)

        cosine_dist = torch.matmul(
            feat_l2,
            F.normalize(self.mapper.weight.T, dim=0)
        )

        sine_dist = torch.sqrt(1 - torch.square(cosine_dist))#.clamp(-1,1)

        out_dist = cosine_dist

        if y!=None :

            margined_cosine_dist = cosine_dist*self.cos_m - sine_dist*self.sin_m # cos(a+b) eqn

            margined_cosine_dist = torch.where(
                cosine_dist>0,
                margined_cosine_dist,
                cosine_dist
            ) # easy margin

            out_dist = margined_cosine_dist*y + cosine_dist*(1-y)

        out = self.s * out_dist

        return 8 * out


class AttentiveResNet50Arc_Classifier(nn.Module):
    def __init__(self, class_count:int):
        super(AttentiveResNet50Arc_Classifier, self).__init__()

        self.feature_network = attentive_resnet50()
        #self.feature_network.conv1 = nn.Conv2d(1, 64, kernel_size=7, stride=2, padding=3, bias=False)
        self.feature_network.fc = EmptyModule()

        self.bn_final = nn.BatchNorm1d(2048)

        self.mapper = nn.Linear(2048, class_count, bias=False)

    def forward(self, x):

        feat = self.feature_network(x)
        out = self.bn_final(feat)
        out = 8 * F.normalize(out, dim=1)

        output = torch.matmul(
            out,
            F.normalize(self.mapper.weight.T, dim=0)
        )

        return output

if __name__=="__main__":
    import torch
    import numpy as np
    import python_speech_features

    audio = np.random.rand(24000)

    mfcc_features = python_speech_features.mfcc(audio,samplerate=16000,winlen=0.025,winstep=0.01,numcep=32,
                     nfilt=32,nfft=512,lowfreq=0,highfreq=None,preemph=0.97,
         ceplifter=64,appendEnergy=True)
    mfcc_features = np.expand_dims(mfcc_features, axis=0)

    net = AttentiveResNet50Arc_Classifier(1750)
    #net = efficientformerv2_s0(pretrained=False, input_size=(1,149,64), input_chan=1)
    #net = timm.create_model('efficientformer_l1', in_chans=1)
    net.eval()
    print(net.training)
    inp = np.expand_dims(mfcc_features, axis=0)
    print(inp.shape)
    out = net(torch.Tensor(inp))
    print(out.shape, inp.shape, mfcc_features.shape)