crowdcount.models

Several crowd counting models implements.

class crowdcount.models.CSRNet(load_weights=False)[source]

Refer from “CSRNet: …” paper.

Parameters

load_weights (bool) – If True, CSRNet will be pre-trained on ImageNet

class crowdcount.models.MCNN(bn=False)[source]

Refer from “MCNN…” paper.

Parameters

bn (bool) – if True, Batch Normalization layer will be used after every convolutional layer

class crowdcount.models.Res50(pretrained=True)[source]

Refer from “C-3-Framework…” paper

Parameters

pretrain (bool) – if True, this model will be pre-trianed on ImageNet

class crowdcount.models.Res101(pretrained=True)[source]

Refer from “C-3-Framework…” paper

Parameters

pretrain (bool) – if True, this model will be pre-trianed on ImageNet

class crowdcount.models.UNet[source]

Refer from “Pytorch-UNet…” paper

Parameters

pretrain (bool) – if True, this model will be pre-trianed on ImageNet

class crowdcount.models.VGG(load_weights=True)[source]

Refer from “VGG…” paper

Parameters

load_weights (bool) – If True, CSRNet will be pre-trained on ImageNet