cs231n assignment3
前一段时间把公开课cs231n看完，然后这里分享下assignment3的代码，水平有限，如有疏漏之处请见谅。assignment3主要内容包括Image Captioning和深度网络可视化。对于Image Captioning，已经预先提取了图像特征不需要自己做，只需要自己实现RNN，LSTM就行，下面是各个作业的代码（写的有点烂，各位凑合着看吧）。
Image Captioning with Vanilla RNNs and Image Captioning with LSTMs
实现Vanilla RNN和lstm
rnn_layers.py
import numpy as np
"""
This file defines layer types that are commonly used for recurrent neural
networks.
"""
def rnn_step_forward(x, prev_h, Wx, Wh, b):
"""
Run the forward pass for a single timestep of a vanilla RNN that uses a tanh
activation function.
The input data has dimension D, the hidden state has dimension H, and we use
a minibatch size of N.
Inputs:
- x: Input data for this timestep, of shape (N, D).
- prev_h: Hidden state from previous timestep, of shape (N, H)
- Wx: Weight matrix for input-to-hidden connections, of shape (D, H)
- Wh: Weight matrix for hidden-to-hidden connections, of shape (H, H)
- b: Biases of shape (H,)
Returns a tuple of:
- next_h: Next hidden state, of shape (N, H)
- cache: Tuple of values needed for the backward pass.
"""
next_h, cache = None, None
##############################################################################
# TODO: Implement a single forward step for the vanilla RNN. Store the next #
# hidden state and any values you need for the backward pass in the next_h #
# and cache variables respectively. #
##############################################################################
N, D = x.shape
_, H = Wx.shape
next_h = np.dot(x, Wx) + np.dot(prev_h, Wh) + b
next_h = np.tanh(next_h)
cache = (x, prev_h, Wx, Wh, next_h)
##############################################################################
# END OF YOUR CODE #
##############################################################################
return next_h, cache
def rnn_step_backward(dnext_h, cache):
"""
Backward pass for a single timestep of a vanilla RNN.
Inputs:
- dnext_h: Gradient of loss with respect to next hidden state
- cache: Cache object from the forward pass
Returns a tuple of:
- dx: Gradients of input data, of shape (N, D)
- dprev_h: Gradients of previous hidden state, of shape (N, H)
- dWx: Gradients of input-to-hidden weights, of shape (D, H)
- dWh: Gradients of hidden-to-hidden weights, of shape (H, H)
- db: Gradients of bias vector, of shape (H,)
"""
dx, dprev_h, dWx, dWh, db = None, None, None, None, None
##############################################################################
# TODO: Implement the backward pass for a single step of a vanilla RNN. #
# #
# HINT: For the tanh function, you can compute the local derivative in terms #
# of the output value from tanh. #
##############################################################################
(x, prev_h, Wx, Wh, next_h) = cache
dtanh = (1 - next_h * next_h) * dnext_h
db = np.sum(dtanh, axis = 0)
dWx = np.dot(x.T, dtanh)
dWh = np.dot(prev_h.T, dtanh)
dprev_h = np.dot(dtanh, Wh.T)
dx = np.dot(dtanh, Wx.T)
##############################################################################
# END OF YOUR CODE #
##############################################################################
return dx, dprev_h, dWx, dWh, db
def rnn_forward(x, h0, Wx, Wh, b):
"""
Run a vanilla RNN forward on an entire sequence of data. We assume an input
sequence composed of T vectors, each of dimension D. The RNN uses a hidden
size of H, and we work over a minibatch containing N sequences. After running
the RNN forward, we return the hidden states for all timesteps.
Inputs:
- x: Input data for the entire timeseries, of shape (N, T, D).
- h0: Initial hidden state, of shape (N, H)
- Wx: Weight matrix for input-to-hidden connections, of shape (D, H)
- Wh: Weight matrix for hidden-to-hidden connections, of shape (H, H)
- b: Biases of shape (H,)
Returns a tuple of:
- h: Hidden states for the entire timeseries, of shape (N, T, H).
- cache: Values needed in the backward pass
"""
h, cache = None, None
##############################################################################
# TODO: Implement forward pass for a vanilla RNN running on a sequence of #
# input data. You should use the rnn_step_forward function that you defined #
# above. #
##############################################################################
N, T, D = x.shape
_, H = Wh.shape
cache = []
h = np.zeros((N, T, H))
for i in np.arange(T):
h0, tmp = rnn_step_forward(x[:,i,:], h0, Wx, Wh, b)
cache.append(tmp)
h[:, i, :] = h0
##############################################################################
# END OF YOUR CODE #
##############################################################################
return h, cache
def rnn_backward(dh, cache):
"""
Compute the backward pass for a vanilla RNN over an entire sequence of data.
Inputs:
- dh: Upstream gradients of all hidden states, of shape (N, T, H)
Returns a tuple of:
- dx: Gradient of inputs, of shape (N, T, D)
- dh0: Gradient of initial hidden state, of shape (N, H)
- dWx: Gradient of input-to-hidden weights, of shape (D, H)
- dWh: Gradient of hidden-to-hidden weights, of shape (H, H)
- db: Gradient of biases, of shape (H,)
"""
dx, dh0, dWx, dWh, db = None, None, None, None, None
##############################################################################
# TODO: Implement the backward pass for a vanilla RNN running an entire #
# sequence of data. You should use the rnn_step_backward function that you #
# defined above. #
##############################################################################
N, T, H = dh.shape
_, _, W, _, _ = cache[0]
D = W.shape[0]
dx = np.zeros((N, T, D))
dh0 = np.zeros((N, H))
dWx = np.zeros((D, H))
dWh = np.zeros((H, H))
db = np.zeros(H)
dh_prev = np.zeros((N, H))
for i in reversed(np.arange(T)):
dh_cur = dh_prev + dh[:, i, :] #copy!!!
dx[:, i, :], dh_prev, tdWx, tdWh, tdb = rnn_step_backward(dh_cur, cache[i])
dWx += tdWx
dWh += tdWh
db += tdb
dh0 = dh_prev
##############################################################################
# END OF YOUR CODE #
##############################################################################
return dx, dh0, dWx, dWh, db
def word_embedding_forward(x, W):
"""
Forward pass for word embeddings. We operate on minibatches of size N where
each sequence has length T. We assume a vocabulary of V words, assigning each
to a vector of dimension D.
Inputs:
- x: Integer array of shape (N, T) giving indices of words. Each element idx
of x muxt be in the range 0 <= idx < V.
- W: Weight matrix of shape (V, D) giving word vectors for all words.
Returns a tuple of:
- out: Array of shape (N, T, D) giving word vectors for all input words.
- cache: Values needed for the backward pass
"""
out, cache = None, None
##############################################################################
# TODO: Implement the forward pass for word embeddings. #
# #
# HINT: This should be very simple. #
##############################################################################
#print x.shape
N, T = x.shape
V, D = W.shape
out = np.zeros((N, T, D))
for i in np.arange(N):
for j in np.arange(T):
out[i, j, :] = W[int(x[i, j]), :]
cache = (x, W)
##############################################################################
# END OF YOUR CODE #
##############################################################################
return out, cache
def word_embedding_backward(dout, cache):
"""
Backward pass for word embeddings. We cannot back-propagate into the words
since they are integers, so we only return gradient for the word embedding
matrix.
HINT: Look up the function np.add.at
Inputs:
- dout: Upstream gradients of shape (N, T, D)
- cache: Values from the forward pass
Returns:
- dW: Gradient of word embedding matrix, of shape (V, D).
"""
dW = None
##############################################################################
# TODO: Implement the backward pass for word embeddings. #
# #
# HINT: Look up the function np.add.at #
##############################################################################
N, T, D = dout.shape
(x, W) = cache
V, _ = W.shape
dW = np.zeros((V, D))
for i in np.arange(N):
for j in np.arange(T):
dW[x[i, j], :] += dout[i, j, :]
##############################################################################
# END OF YOUR CODE #
##############################################################################
return dW
def sigmoid(x):
"""
A numerically stable version of the logistic sigmoid function.
"""
pos_mask = (x >= 0)
neg_mask = (x < 0)
z = np.zeros_like(x)
z[pos_mask] = np.exp(-x[pos_mask])
z[neg_mask] = np.exp(x[neg_mask])
top = np.ones_like(x)
top[neg_mask] = z[neg_mask]
return top / (1 + z)
def lstm_step_forward(x, prev_h, prev_c, Wx, Wh, b):
"""
Forward pass for a single timestep of an LSTM.
The input data has dimension D, the hidden state has dimension H, and we use
a minibatch size of N.
Inputs:
- x: Input data, of shape (N, D)
- prev_h: Previous hidden state, of shape (N, H)
- prev_c: previous cell state, of shape (N, H)
- Wx: Input-to-hidden weights, of shape (D, 4H)
- Wh: Hidden-to-hidden weights, of shape (H, 4H)
- b: Biases, of shape (4H,)
Returns a tuple of:
- next_h: Next hidden state, of shape (N, H)
- next_c: Next cell state, of shape (N, H)
- cache: Tuple of values needed for backward pass.
"""
next_h, next_c, cache = None, None, None
#############################################################################
# TODO: Implement the forward pass for a single timestep of an LSTM. #
# You may want to use the numerically stable sigmoid implementation above. #
#############################################################################
N, D = x.shape
_, H = prev_h.shape
a = np.dot(x, Wx) + np.dot(prev_h, Wh) + b
i = sigmoid(a[:, :H])
f = sigmoid(a[:, H:2*H])
o = sigmoid(a[:, 2*H:3*H])
g = np.tanh(a[:, 3*H:])
next_c = f * prev_c + i * g
tanh_c = np.tanh(next_c)
next_h = o * tanh_c
cache = (a, i, f, o, g, tanh_c, Wx, Wh, b, x, prev_c, prev_h)
##############################################################################
# END OF YOUR CODE #
##############################################################################
return next_h, next_c, cache
def lstm_step_backward(dnext_h, dnext_c, cache):
"""
Backward pass for a single timestep of an LSTM.
Inputs:
- dnext_h: Gradients of next hidden state, of shape (N, H)
- dnext_c: Gradients of next cell state, of shape (N, H)
- cache: Values from the forward pass
Returns a tuple of:
- dx: Gradient of input data, of shape (N, D)
- dprev_h: Gradient of previous hidden state, of shape (N, H)
- dprev_c: Gradient of previous cell state, of shape (N, H)
- dWx: Gradient of input-to-hidden weights, of shape (D, 4H)
- dWh: Gradient of hidden-to-hidden weights, of shape (H, 4H)
- db: Gradient of biases, of shape (4H,)
"""
dx, dprev_h, dprev_c, dWx, dWh, db = None, None, None, None, None, None
#############################################################################
# TODO: Implement the backward pass for a single timestep of an LSTM. #
# #
# HINT: For sigmoid and tanh you can compute local derivatives in terms of #
# the output value from the nonlinearity. #
#############################################################################
(a, i, f, o, g, tanh_c, Wx, Wh, b, x, prev_c, prev_h) = cache
N, D = x.shape
_, H = prev_h.shape
dx = np.zeros_like(a)
do = dnext_h * tanh_c # N x H
dtanh_c = dnext_h * o # N x H
dc = dtanh_c * (1 - tanh_c * tanh_c) # N x H
#print dc.shape
#print dnext_c.shape
dc = dc + dnext_c # N x H ??????
df = dc * prev_c # N x H
dprev_c = dc * f # N x H
di = dc * g # N x H
dg = dc * i # N x H
da = np.zeros_like(a)
da[:, :H] = di * (1 - i) * i
da[:, H:2*H] = df * (1 - f) * f
da[:, 2*H:3*H] = do * (1 - o) * o
da[:, 3*H:] = dg * (1 - g * g )
#da = np.array([da_1,da_2:da_3:da_4]) # N x 4H
#print da.shape
db = np.sum(da, axis = 0)
dWx = np.dot(x.T, da )
dWh = np.dot(prev_h.T, da)
dx = np.dot(da, Wx.T)
dprev_h = np.dot(da, Wh.T)
##############################################################################
# END OF YOUR CODE #
##############################################################################
return dx, dprev_h, dprev_c, dWx, dWh, db
def lstm_forward(x, h0, Wx, Wh, b):
"""
Forward pass for an LSTM over an entire sequence of data. We assume an input
sequence composed of T vectors, each of dimension D. The LSTM uses a hidden
size of H, and we work over a minibatch containing N sequences. After running
the LSTM forward, we return the hidden states for all timesteps.
Note that the initial cell state is passed as input, but the initial cell
state is set to zero. Also note that the cell state is not returned; it is
an internal variable to the LSTM and is not accessed from outside.
Inputs:
- x: Input data of shape (N, T, D)
- h0: Initial hidden state of shape (N, H)
- Wx: Weights for input-to-hidden connections, of shape (D, 4H)
- Wh: Weights for hidden-to-hidden connections, of shape (H, 4H)
- b: Biases of shape (4H,)
Returns a tuple of:
- h: Hidden states for all timesteps of all sequences, of shape (N, T, H)
- cache: Values needed for the backward pass.
"""
h, cache = None, None
#############################################################################
# TODO: Implement the forward pass for an LSTM over an entire timeseries. #
# You should use the lstm_step_forward function that you just defined. #
#############################################################################
N, T, D = x.shape
_, H = h0.shape
cache = []
h = np.zeros((N, T, H))
c0 = np.zeros((N, H))
for i in np.arange(T):
h0, c0, tmp = lstm_step_forward(x[:, i, :], h0, c0, Wx, Wh, b)
h[:, i, :] = h0
cache.append(tmp)
##############################################################################
# END OF YOUR CODE #
##############################################################################
return h, cache
def lstm_backward(dh, cache):
"""
Backward pass for an LSTM over an entire sequence of data.]
Inputs:
- dh: Upstream gradients of hidden states, of shape (N, T, H)
- cache: Values from the forward pass
Returns a tuple of:
- dx: Gradient of input data of shape (N, T, D)
- dh0: Gradient of initial hidden state of shape (N, H)
- dWx: Gradient of input-to-hidden weight matrix of shape (D, 4H)
- dWh: Gradient of hidden-to-hidden weight matrix of shape (H, 4H)
- db: Gradient of biases, of shape (4H,)
"""
dx, dh0, dWx, dWh, db = None, None, None, None, None
#############################################################################
# TODO: Implement the backward pass for an LSTM over an entire timeseries. #
# You should use the lstm_step_backward function that you just defined. #
#############################################################################
N, T, H = dh.shape
D = cache[0][6].shape[0]
dc = np.zeros((N, H))
dx = np.zeros((N, T, D))
dWx = np.zeros((D, 4*H))
dWh = np.zeros((H, 4*H))
db = np.zeros(4*H)
dh0 = np.zeros((N, H))
for i in reversed(np.arange(T)):
dh0 += dh[:, i, :]
dx[:, i, :], dh0, dc, dw, dwh, tdb = lstm_step_backward(dh0, dc, cache[i])
dWx += dw
dWh += dwh
#print tdb.shape
#print db.shape
db += tdb
##############################################################################
# END OF YOUR CODE #
##############################################################################
return dx, dh0, dWx, dWh, db
def temporal_affine_forward(x, w, b):
"""
Forward pass for a temporal affine layer. The input is a set of D-dimensional
vectors arranged into a minibatch of N timeseries, each of length T. We use
an affine function to transform each of those vectors into a new vector of
dimension M.
Inputs:
- x: Input data of shape (N, T, D)
- w: Weights of shape (D, M)
- b: Biases of shape (M,)
Returns a tuple of:
- out: Output data of shape (N, T, M)
- cache: Values needed for the backward pass
"""
N, T, D = x.shape
M = b.shape[0]
out = x.reshape(N * T, D).dot(w).reshape(N, T, M) + b
cache = x, w, b, out
return out, cache
def temporal_affine_backward(dout, cache):
"""
Backward pass for temporal affine layer.
Input:
- dout: Upstream gradients of shape (N, T, M)
- cache: Values from forward pass
Returns a tuple of:
- dx: Gradient of input, of shape (N, T, D)
- dw: Gradient of weights, of shape (D, M)
- db: Gradient of biases, of shape (M,)
"""
x, w, b, out = cache
N, T, D = x.shape
M = b.shape[0]
dx = dout.reshape(N * T, M).dot(w.T).reshape(N, T, D)
dw = dout.reshape(N * T, M).T.dot(x.reshape(N * T, D)).T
db = dout.sum(axis=(0, 1))
return dx, dw, db
def temporal_softmax_loss(x, y, mask, verbose=False):
"""
A temporal version of softmax loss for use in RNNs. We assume that we are
making predictions over a vocabulary of size V for each timestep of a
timeseries of length T, over a minibatch of size N. The input x gives scores
for all vocabulary elements at all timesteps, and y gives the indices of the
ground-truth element at each timestep. We use a cross-entropy loss at each
timestep, summing the loss over all timesteps and averaging across the
minibatch.
As an additional complication, we may want to ignore the model output at some
timesteps, since sequences of different length may have been combined into a
minibatch and padded with NULL tokens. The optional mask argument tells us
which elements should contribute to the loss.
Inputs:
- x: Input scores, of shape (N, T, V)
- y: Ground-truth indices, of shape (N, T) where each element is in the range
0 <= y[i, t] < V
- mask: Boolean array of shape (N, T) where mask[i, t] tells whether or not
the scores at x[i, t] should contribute to the loss.
Returns a tuple of:
- loss: Scalar giving loss
- dx: Gradient of loss with respect to scores x.
"""
N, T, V = x.shape
x_flat = x.reshape(N * T, V)
y_flat = y.reshape(N * T)
mask_flat = mask.reshape(N * T)
probs = np.exp(x_flat - np.max(x_flat, axis=1, keepdims=True))
probs /= np.sum(probs, axis=1, keepdims=True)
loss = -np.sum(mask_flat * np.log(probs[np.arange(N * T), y_flat])) / N
dx_flat = probs.copy()
dx_flat[np.arange(N * T), y_flat] -= 1
dx_flat /= N
dx_flat *= mask_flat[:, None]
if verbose: print 'dx_flat: ', dx_flat.shape
dx = dx_flat.reshape(N, T, V)
return loss, dx
rnn.py
import numpy as np
from cs231n.layers import *
from cs231n.rnn_layers import *
class CaptioningRNN(object):
"""
A CaptioningRNN produces captions from image features using a recurrent
neural network.
The RNN receives input vectors of size D, has a vocab size of V, works on
sequences of length T, has an RNN hidden dimension of H, uses word vectors
of dimension W, and operates on minibatches of size N.
Note that we don't use any regularization for the CaptioningRNN.
"""
def __init__(self, word_to_idx, input_dim=512, wordvec_dim=128,
hidden_dim=128, cell_type='rnn', dtype=np.float32):
"""
Construct a new CaptioningRNN instance.
Inputs:
- word_to_idx: A dictionary giving the vocabulary. It contains V entries,
and maps each string to a unique integer in the range [0, V).
- input_dim: Dimension D of input image feature vectors.
- wordvec_dim: Dimension W of word vectors.
- hidden_dim: Dimension H for the hidden state of the RNN.
- cell_type: What type of RNN to use; either 'rnn' or 'lstm'.
- dtype: numpy datatype to use; use float32 for training and float64 for
numeric gradient checking.
"""
if cell_type not in {'rnn', 'lstm'}:
raise ValueError('Invalid cell_type "%s"' % cell_type)
self.cell_type = cell_type
self.dtype = dtype
self.word_to_idx = word_to_idx
self.idx_to_word = {i: w for w, i in word_to_idx.iteritems()}
self.params = {}
vocab_size = len(word_to_idx)
self._null = word_to_idx['' ]
self._start = word_to_idx.get('' , None)
self._end = word_to_idx.get('' , None)
# Initialize word vectors
self.params['W_embed'] = np.random.randn(vocab_size, wordvec_dim)
self.params['W_embed'] /= 100
# Initialize CNN -> hidden state projection parameters
self.params['W_proj'] = np.random.randn(input_dim, hidden_dim)
self.params['W_proj'] /= np.sqrt(input_dim)
self.params['b_proj'] = np.zeros(hidden_dim)
# Initialize parameters for the RNN
dim_mul = {'lstm': 4, 'rnn': 1}[cell_type]
self.params['Wx'] = np.random.randn(wordvec_dim, dim_mul * hidden_dim)
self.params['Wx'] /= np.sqrt(wordvec_dim)
self.params['Wh'] = np.random.randn(hidden_dim, dim_mul * hidden_dim)
self.params['Wh'] /= np.sqrt(hidden_dim)
self.params['b'] = np.zeros(dim_mul * hidden_dim)
# Initialize output to vocab weights
self.params['W_vocab'] = np.random.randn(hidden_dim, vocab_size)
self.params['W_vocab'] /= np.sqrt(hidden_dim)
self.params['b_vocab'] = np.zeros(vocab_size)
# Cast parameters to correct dtype
for k, v in self.params.iteritems():
self.params[k] = v.astype(self.dtype)
def loss(self, features, captions):
"""
Compute training-time loss for the RNN. We input image features and
ground-truth captions for those images, and use an RNN (or LSTM) to compute
loss and gradients on all parameters.
Inputs:
- features: Input image features, of shape (N, D)
- captions: Ground-truth captions; an integer array of shape (N, T) where
each element is in the range 0 <= y[i, t] < V
Returns a tuple of:
- loss: Scalar loss
- grads: Dictionary of gradients parallel to self.params
"""
# Cut captions into two pieces: captions_in has everything but the last word
# and will be input to the RNN; captions_out has everything but the first
# word and this is what we will expect the RNN to generate. These are offset
# by one relative to each other because the RNN should produce word (t+1)
# after receiving word t. The first element of captions_in will be the START
# token, and the first element of captions_out will be the first word.
captions_in = captions[:, :-1]
captions_out = captions[:, 1:]
# You'll need this
mask = (captions_out != self._null)
# Weight and bias for the affine transform from image features to initial
# hidden state
W_proj, b_proj = self.params['W_proj'], self.params['b_proj']
# Word embedding matrix
W_embed = self.params['W_embed']
# Input-to-hidden, hidden-to-hidden, and biases for the RNN
Wx, Wh, b = self.params['Wx'], self.params['Wh'], self.params['b']
# Weight and bias for the hidden-to-vocab transformation.
W_vocab, b_vocab = self.params['W_vocab'], self.params['b_vocab']
loss, grads = 0.0, {}
############################################################################
# TODO: Implement the forward and backward passes for the CaptioningRNN. #
# In the forward pass you will need to do the following: #
# (1) Use an affine transformation to compute the initial hidden state #
# from the image features. This should produce an array of shape (N, H)#
# (2) Use a word embedding layer to transform the words in captions_in #
# from indices to vectors, giving an array of shape (N, T, W). #
# (3) Use either a vanilla RNN or LSTM (depending on self.cell_type) to #
# process the sequence of input word vectors and produce hidden state #
# vectors for all timesteps, producing an array of shape (N, T, H). #
# (4) Use a (temporal) affine transformation to compute scores over the #
# vocabulary at every timestep using the hidden states, giving an #
# array of shape (N, T, V). #
# (5) Use (temporal) softmax to compute loss using captions_out, ignoring #
# the points where the output word is using the mask above. #
# #
# In the backward pass you will need to compute the gradient of the loss #
# with respect to all model parameters. Use the loss and grads variables #
# defined above to store loss and gradients; grads[k] should give the #
# gradients for self.params[k]. #
############################################################################
"""
forward pass
"""
#first layer affine transformation to calculate h0
h0, flcache = affine_forward(features, W_proj, b_proj)
# Using embedding layer to get word vectors
fx, emcache = word_embedding_forward(captions_in, W_embed)
# RNN or LSTM
if(self.cell_type == 'rnn'):
fx, hicache = rnn_forward(fx, h0, Wx, Wh, b)
else: #if(self.cell_type == 'lstm'):
fx, hicache = lstm_forward(fx, h0, Wx, Wh, b)
# affine transformation
fx, affcache = temporal_affine_forward(fx, W_vocab, b_vocab)
# softmax
loss, dx = temporal_softmax_loss(fx, captions_out, mask)
"""
backward pass
"""
# affine trans backward
dx, grads['W_vocab'], grads['b_vocab'] = temporal_affine_backward(dx, affcache)
# RNN or LSTM backward
if(self.cell_type == 'rnn'):
dx, dh0, grads['Wx'], grads['Wh'], grads['b'] = rnn_backward(dx, hicache)
else:
dx, dh0, grads['Wx'], grads['Wh'], grads['b'] = lstm_backward(dx, hicache)
# embedding backward
grads['W_embed'] = word_embedding_backward(dx, emcache)
# first layer(affine transform) backward
_, grads['W_proj'], grads['b_proj'] = affine_backward(dh0, flcache)
############################################################################
# END OF YOUR CODE #
############################################################################
return loss, grads
def sample(self, features, max_length=30):
"""
Run a test-time forward pass for the model, sampling captions for input
feature vectors.
At each timestep, we embed the current word, pass it and the previous hidden
state to the RNN to get the next hidden state, use the hidden state to get
scores for all vocab words, and choose the word with the highest score as
the next word. The initial hidden state is computed by applying an affine
transform to the input image features, and the initial word is the
token.
For LSTMs you will also have to keep track of the cell state; in that case
the initial cell state should be zero.
Inputs:
- features: Array of input image features of shape (N, D).
- max_length: Maximum length T of generated captions.
Returns:
- captions: Array of shape (N, max_length) giving sampled captions,
where each element is an integer in the range [0, V). The first element
of captions should be the first sampled word, not the token.
"""
N = features.shape[0]
captions = self._null * np.ones((N, max_length + 1), dtype=np.int32)
# Unpack parameters
W_proj, b_proj = self.params['W_proj'], self.params['b_proj']
W_embed = self.params['W_embed']
Wx, Wh, b = self.params['Wx'], self.params['Wh'], self.params['b']
W_vocab, b_vocab = self.params['W_vocab'], self.params['b_vocab']
###########################################################################
# TODO: Implement test-time sampling for the model. You will need to #
# initialize the hidden state of the RNN by applying the learned affine #
# transform to the input image features. The first word that you feed to #
# the RNN should be the token; its value is stored in the #
# variable self._start. At each timestep you will need to do to: #
# (1) Embed the previous word using the learned word embeddings #
# (2) Make an RNN step using the previous hidden state and the embedded #
# current word to get the next hidden state. #
# (3) Apply the learned affine transformation to the next hidden state to #
# get scores for all words in the vocabulary #
# (4) Select the word with the highest score as the next word, writing it #
# to the appropriate slot in the captions variable #
# #
# For simplicity, you do not need to stop generating after an token #
# is sampled, but you can if you want to. #
# #
# HINT: You will not be able to use the rnn_forward or lstm_forward #
# functions; you'll need to call rnn_step_forward or lstm_step_forward in #
# a loop. #
###########################################################################
"""
sampling process
"""
# get h0
h0, _ = affine_forward(features, W_proj, b_proj)
# generate captions
nc = np.zeros_like(h0)
word = np.ones((N, 1)) * self._start
captions[:, 0] = self._start
for i in np.arange(max_length):
fx, _ = word_embedding_forward(word, W_embed)
#print fx.shape
if( self.cell_type == 'rnn' ):
h0, _ = rnn_step_forward(np.squeeze(fx), h0, Wx, Wh, b)
else:
h0, nc, _ = lstm_step_forward( np.squeeze(fx), h0, nc, Wx, Wh, b )
#print fx.shape
#h0 = np.squeeze(fx)
#h0 = copy(fx)
fx, _ = temporal_affine_forward(h0[:, np.newaxis,:], W_vocab, b_vocab) # copy
_, T, D = fx.shape
#print fx.shape
word = np.argmax(fx, axis = 2)
#word = word.reshape(-1, 1)
#word = self.idx_to_word[word_idx]
#print word.shape
captions[:, i + 1] = np.squeeze(word)
############################################################################
# END OF YOUR CODE #
############################################################################
return captionsImage Gradients: Saliency maps and Fooling Images
Saliency Maps
from cs231n.layers import svm_loss
def compute_saliency_maps(X, y, model):
"""
Compute a class saliency map using the model for images X and labels y.
Input:
- X: Input images, of shape (N, 3, H, W)
- y: Labels for X, of shape (N,)
- model: A PretrainedCNN that will be used to compute the saliency map.
Returns:
- saliency: An array of shape (N, H, W) giving the saliency maps for the input
images.
"""
print X.shape
saliency = None
##############################################################################
# TODO: Implement this function. You should use the forward and backward #
# methods of the PretrainedCNN class, and compute gradients with respect to #
# the unnormalized class score of the ground-truth classes in y. #
##############################################################################
out, cache = model.forward(X, None, None, 'test')
_, dsc = svm_loss(out, y)
print dsc.shape
dx = np.zeros_like(dsc)
dx[np.arange(X.shape[0]), y] = 1 #better than svm_loss dx
dx, _ = model.backward(dx, cache)
saliency = np.amax(np.absolute(dx), axis = 1)
#print saliency.shape
#print X.shape
##############################################################################
# END OF YOUR CODE #
##############################################################################
return saliencyFooling Images
from cs231n.layers import softmax_loss
def make_fooling_image(X, target_y, model):
"""
Generate a fooling image that is close to X, but that the model classifies
as target_y.
Inputs:
- X: Input image, of shape (1, 3, 64, 64)
- target_y: An integer in the range [0, 100)
- model: A PretrainedCNN
Returns:
- X_fooling: An image that is close to X, but that is classifed as target_y
by the model.
"""
X_fooling = X.copy()
##############################################################################
# TODO: Generate a fooling image X_fooling that the model will classify as #
# the class target_y. Use gradient ascent on the target class score, using #
# the model.forward method to compute scores and the model.backward method #
# to compute image gradients. #
# #
# HINT: For most examples, you should be able to generate a fooling image #
# in fewer than 100 iterations of gradient ascent. #
##############################################################################
i = 1
while True:
out, cache = model.forward(X_fooling, None, None, 'test')
if( np.argmax(out) == target_y):
print i
break
_, dx = svm_loss(out, target_y)
#dx= np.zeros_like(out)
#dx[np.arange(X.shape[0]),target_y] = 1.0
dx, _ = model.backward(dx, cache)
lr = 100
X_fooling -= 100 * dx #using svm_loss should be - (gradient descent)
if( i > 100):
print np.argmax(out)
print "Eoor!"
break
i += 1
##############################################################################
# END OF YOUR CODE #
##############################################################################
return X_foolingImage Generation: Classes, Inversion, DeepDream
Class visualization
def create_class_visualization(target_y, model, **kwargs):
"""
Perform optimization over the image to generate class visualizations.
Inputs:
- target_y: Integer in the range [0, 100) giving the target class
- model: A PretrainedCNN that will be used for generation
Keyword arguments:
- learning_rate: Floating point number giving the learning rate
- blur_every: An integer; how often to blur the image as a regularizer
- l2_reg: Floating point number giving L2 regularization strength on the image;
this is lambda in the equation above.
- max_jitter: How much random jitter to add to the image as regularization
- num_iterations: How many iterations to run for
- show_every: How often to show the image
"""
learning_rate = kwargs.pop('learning_rate', 10000)
blur_every = kwargs.pop('blur_every', 1)
l2_reg = kwargs.pop('l2_reg', 1e-6)
max_jitter = kwargs.pop('max_jitter', 4)
num_iterations = kwargs.pop('num_iterations', 100)
show_every = kwargs.pop('show_every', 25)
X = np.random.randn(1, 3, 64, 64)
for t in xrange(num_iterations):
# As a regularizer, add random jitter to the image
ox, oy = np.random.randint(-max_jitter, max_jitter+1, 2)
X = np.roll(np.roll(X, ox, -1), oy, -2)
dX = None
############################################################################
# TODO: Compute the image gradient dX of the image with respect to the #
# target_y class score. This should be similar to the fooling images. Also #
# add L2 regularization to dX and update the image X using the image #
# gradient and the learning rate. #
############################################################################
out, cache = model.forward(X, mode = 'test')
dout = np.zeros_like(out)
dout[np.arange(X.shape[0]), target_y] = 1.0
dx, _ = model.backward(dout, cache)
dx += 2 * l2_reg * X
X += learning_rate * dx
############################################################################
# END OF YOUR CODE #
############################################################################
# Undo the jitter
X = np.roll(np.roll(X, -ox, -1), -oy, -2)
# As a regularizer, clip the image
X = np.clip(X, -data['mean_image'], 255.0 - data['mean_image'])
# As a regularizer, periodically blur the image
if t % blur_every == 0:
X = blur_image(X)
# Periodically show the image
if t % show_every == 0:
plt.imshow(deprocess_image(X, data['mean_image']))
plt.gcf().set_size_inches(3, 3)
plt.axis('off')
plt.show()
return XFeature Inversion
def invert_features(target_feats, layer, model, **kwargs):
"""
Perform feature inversion in the style of Mahendran and Vedaldi 2015, using
L2 regularization and periodic blurring.
Inputs:
- target_feats: Image features of the target image, of shape (1, C, H, W);
we will try to generate an image that matches these features
- layer: The index of the layer from which the features were extracted
- model: A PretrainedCNN that was used to extract features
Keyword arguments:
- learning_rate: The learning rate to use for gradient descent
- num_iterations: The number of iterations to use for gradient descent
- l2_reg: The strength of L2 regularization to use; this is lambda in the
equation above.
- blur_every: How often to blur the image as implicit regularization; set
to 0 to disable blurring.
- show_every: How often to show the generated image; set to 0 to disable
showing intermediate reuslts.
Returns:
- X: Generated image of shape (1, 3, 64, 64) that matches the target features.
"""
learning_rate = kwargs.pop('learning_rate', 10000)
num_iterations = kwargs.pop('num_iterations', 500)
l2_reg = kwargs.pop('l2_reg', 1e-7)
blur_every = kwargs.pop('blur_every', 1)
show_every = kwargs.pop('show_every', 50)
X = np.random.randn(1, 3, 64, 64)
for t in xrange(num_iterations):
############################################################################
# TODO: Compute the image gradient dX of the reconstruction loss with #
# respect to the image. You should include L2 regularization penalizing #
# large pixel values in the generated image using the l2_reg parameter; #
# then update the generated image using the learning_rate from above. #
############################################################################
out, cache = model.forward(X, end = layer, mode = 'test')
loss = np.sum((out - target_feats)**2) + l2_reg * np.sum(X**2)
dout = 2 * (out - target_feats)
dx, _ = model.backward(dout, cache)
dx += 2 * l2_reg * X
X -= learning_rate * dx
############################################################################
# END OF YOUR CODE #
############################################################################
# As a regularizer, clip the image
X = np.clip(X, -data['mean_image'], 255.0 - data['mean_image'])
# As a regularizer, periodically blur the image
if (blur_every > 0) and t % blur_every == 0:
X = blur_image(X)
if (show_every > 0) and (t % show_every == 0 or t + 1 == num_iterations):
print 'loss: %f' %loss
plt.imshow(deprocess_image(X, data['mean_image']))
plt.gcf().set_size_inches(3, 3)
plt.axis('off')
plt.title('t = %d' % t)
plt.show()DeepDream
def deepdream(X, layer, model, **kwargs):
"""
Generate a DeepDream image.
Inputs:
- X: Starting image, of shape (1, 3, H, W)
- layer: Index of layer at which to dream
- model: A PretrainedCNN object
Keyword arguments:
- learning_rate: How much to update the image at each iteration
- max_jitter: Maximum number of pixels for jitter regularization
- num_iterations: How many iterations to run for
- show_every: How often to show the generated image
"""
X = X.copy()
learning_rate = kwargs.pop('learning_rate', 5.0)
max_jitter = kwargs.pop('max_jitter', 16)
num_iterations = kwargs.pop('num_iterations', 100)
show_every = kwargs.pop('show_every', 25)
for t in xrange(num_iterations):
# As a regularizer, add random jitter to the image
ox, oy = np.random.randint(-max_jitter, max_jitter+1, 2)
X = np.roll(np.roll(X, ox, -1), oy, -2)
dX = None
############################################################################
# TODO: Compute the image gradient dX using the DeepDream method. You'll #
# need to use the forward and backward methods of the model object to #
# extract activations and set gradients for the chosen layer. After #
# computing the image gradient dX, you should use the learning rate to #
# update the image X. #
############################################################################
out, cache = model.forward(X, end = layer, mode = 'test')
dx, _ = model.backward(out, cache)
X += learning_rate * dx
############################################################################
# END OF YOUR CODE #
############################################################################
# Undo the jitter
X = np.roll(np.roll(X, -ox, -1), -oy, -2)
# As a regularizer, clip the image
mean_pixel = data['mean_image'].mean(axis=(1, 2), keepdims=True)
X = np.clip(X, -mean_pixel, 255.0 - mean_pixel)
# Periodically show the image
if t == 0 or (t + 1) % show_every == 0:
img = deprocess_image(X, data['mean_image'], mean='pixel')
plt.imshow(img)
plt.title('t = %d' % (t + 1))
plt.gcf().set_size_inches(8, 8)
plt.axis('off')
plt.show()
return X