(self, **kwargs)
| 60 | self.classifier = nn.Linear(fea_dim,2) |
| 61 | |
| 62 | def forward(self, **kwargs): |
| 63 | frames=kwargs['frames'] |
| 64 | fea_img = self.linear_img(frames) |
| 65 | fea_img = self.attention(fea_img) |
| 66 | fea_img = torch.mean(fea_img, -2) |
| 67 | output = self.classifier(fea_img) |
| 68 | return output |
| 69 | |
| 70 | class bVggish(torch.nn.Module): |
| 71 | def __init__(self,fea_dim): |
nothing calls this directly
no outgoing calls
no test coverage detected