natanielruiz
2017-09-21 e692eed7bff06eae5c123350eee21adfa06653e2
code/hopenet.py
@@ -1,8 +1,8 @@
import torch
import torch.nn as nn
import torchvision.datasets as dsets
from torch.autograd import Variable
import math
import torch.nn.functional as F
# CNN Model (2 conv layer)
class Simple_CNN(nn.Module):
@@ -41,7 +41,7 @@
class Hopenet(nn.Module):
    # This is just Hopenet with 3 output layers for yaw, pitch and roll.
    def __init__(self, block, layers, num_bins):
    def __init__(self, block, layers, num_bins, iter_ref):
        self.inplanes = 64
        super(Hopenet, self).__init__()
        self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3,
@@ -57,6 +57,13 @@
        self.fc_yaw = nn.Linear(512 * block.expansion, num_bins)
        self.fc_pitch = nn.Linear(512 * block.expansion, num_bins)
        self.fc_roll = nn.Linear(512 * block.expansion, num_bins)
        self.softmax = nn.Softmax()
        self.fc_finetune = nn.Linear(512 * block.expansion + 3, 3)
        self.idx_tensor = Variable(torch.FloatTensor(range(66))).cuda()
        self.iter_ref = iter_ref
        for m in self.modules():
            if isinstance(m, nn.Conv2d):
@@ -96,8 +103,25 @@
        x = self.avgpool(x)
        x = x.view(x.size(0), -1)
        yaw = self.fc_yaw(x)
        pitch = self.fc_pitch(x)
        roll = self.fc_roll(x)
        pre_yaw = self.fc_yaw(x)
        pre_pitch = self.fc_pitch(x)
        pre_roll = self.fc_roll(x)
        return yaw, pitch, roll
        yaw = self.softmax(pre_yaw)
        yaw = Variable(torch.sum(yaw.data * self.idx_tensor.data, 1), requires_grad=True)
        pitch = self.softmax(pre_pitch)
        pitch = Variable(torch.sum(pitch.data * self.idx_tensor.data, 1), requires_grad=True)
        roll = self.softmax(pre_roll)
        roll = Variable(torch.sum(roll.data * self.idx_tensor.data, 1), requires_grad=True)
        yaw = yaw.view(yaw.size(0), 1)
        pitch = pitch.view(pitch.size(0), 1)
        roll = roll.view(roll.size(0), 1)
        angles = []
        preangles = torch.cat([yaw, pitch, roll], 1)
        angles.append(preangles)
        # angles predicts the residual
        for idx in xrange(self.iter_ref):
            angles.append(self.fc_finetune(torch.cat((preangles, x), 1)))
        return pre_yaw, pre_pitch, pre_roll, angles