fix some bugs

Author: yunlongdong

infCNN/LayerLib.py

@@ -19,8 +19,8 @@ def forward(self, x):
         return y.T + self.bias.reshape((1, -1))
 
     def load_from_torch(self, weights, bias):
-        self.K = weights
-        self.bias = bias
+        self.K = weights.copy()
+        self.bias = bias.copy()
 
 
 class Conv2d(Layer):
@@ -45,11 +45,13 @@ def __init__(self, C_in, C_out, K_s, Stride):
 
     def forward(self, X):
         out = img2col(X, self.K)
-        print(out.shape)
+        out = out.transpose(0, 2, 3, 1)
+        out = out + self.bias
+        return out.transpose(0, 3, 1, 2)
 
     def load_from_torch(self, weights, bias):
-        self.K = weights
-        self.bias = bias
+        self.K = weights.copy()
+        self.bias = bias.copy()
 
 
 

infCNN/OpLib.py

@@ -1,6 +1,6 @@
 from .Core import Op 
-from .im2col import im2col_indices
 import numpy as np
+from .im2col import im2col_indices 
 
 
 class ReLU(Op):

infCNN/conv.py

@@ -34,7 +34,8 @@ def img2col(img, core, buf=[np.zeros(1, dtype=np.int32)]):
 
     # mark where need
     iimg = img.view(dtype=np.int32)
-    iimg.view(dtype=np.uint8).ravel()[::4]&=0xfe
+    # iimg.view(dtype=np.uint8).ravel()[::4]&=0xfe
+    iimg &= 0xfffffffe
     iimg[:,0,:,:] |= 1
 
     # ravel the image
@@ -60,7 +61,8 @@ def img2col(img, core, buf=[np.zeros(1, dtype=np.int32)]):
     img.ravel()[:] = np.arange(3*512*512)
 
     iimg = img.view(dtype=np.int32)
-    iimg.view(dtype=np.uint8).ravel()[::4]&=0xfe
+    # iimg.view(dtype=np.uint8).ravel()[::4]&=0xfe
+    iimg &= 0xfffffffe
     iimg[:,0,:,:] |= 1
 
     core = np.zeros((32, 3, 3, 3), dtype=np.float32)

infCNN/im2col.py

@@ -0,0 +1,51 @@
+import numpy as np
+
+
+def get_im2col_indices(x_shape, field_height, field_width, padding=1, stride=1):
+    # First figure out what the size of the output should be
+    N, C, H, W = x_shape
+    assert (H + 2 * padding - field_height) % stride == 0
+    assert (W + 2 * padding - field_height) % stride == 0
+    out_height = int((H + 2 * padding - field_height) / stride + 1)
+    out_width = int((W + 2 * padding - field_width) / stride + 1)
+
+    i0 = np.repeat(np.arange(field_height), field_width)
+    i0 = np.tile(i0, C)
+    i1 = stride * np.repeat(np.arange(out_height), out_width)
+    j0 = np.tile(np.arange(field_width), field_height * C)
+    j1 = stride * np.tile(np.arange(out_width), out_height)
+    i = i0.reshape(-1, 1) + i1.reshape(1, -1)
+    j = j0.reshape(-1, 1) + j1.reshape(1, -1)
+
+    k = np.repeat(np.arange(C), field_height * field_width).reshape(-1, 1)
+
+    return (k.astype(int), i.astype(int), j.astype(int))
+
+
+def im2col_indices(x, field_height, field_width, padding=1, stride=1):
+    """ An implementation of im2col based on some fancy indexing """
+    # Zero-pad the input
+    p = padding
+    x_padded = np.pad(x, ((0, 0), (0, 0), (p, p), (p, p)), mode='constant')
+
+    k, i, j = get_im2col_indices(x.shape, field_height, field_width, padding, stride)
+
+    cols = x_padded[:, k, i, j]
+    C = x.shape[1]
+    cols = cols.transpose(1, 2, 0).reshape(field_height * field_width * C, -1)
+    return cols
+
+
+def col2im_indices(cols, x_shape, field_height=3, field_width=3, padding=1,
+                   stride=1):
+    """ An implementation of col2im based on fancy indexing and np.add.at """
+    N, C, H, W = x_shape
+    H_padded, W_padded = H + 2 * padding, W + 2 * padding
+    x_padded = np.zeros((N, C, H_padded, W_padded), dtype=cols.dtype)
+    k, i, j = get_im2col_indices(x_shape, field_height, field_width, padding, stride)
+    cols_reshaped = cols.reshape(C * field_height * field_width, -1, N)
+    cols_reshaped = cols_reshaped.transpose(2, 0, 1)
+    np.add.at(x_padded, (slice(None), k, i, j), cols_reshaped)
+    if padding == 0:
+        return x_padded
+    return x_padded[:, :, padding:-padding, padding:-padding]

net.py

@@ -52,9 +52,12 @@ def load_mat(self, fn):
 
 if __name__ == "__main__":
     net = LeNet()
-    data =  np.ones((1, 1, 28, 28))
-    net.load_mat('train/lenet.mat')
+    data =  np.ones((1, 1, 28, 28), dtype='float32')
+    x = np.ascontiguousarray(data, dtype=np.float32)
+    print(data.shape, data.dtype)
+    # net.load_mat('train/lenet.mat')
+    print(net(x))
     start = time()
-    print(net(data))
+    print(net(x))
     print('infCNN LeNet time:', time()-start)