PaddleFluid简介
前面已经介绍了如何在PaddlePaddle下做图像分类工作，以及搭配VisualDL做metric可视化。今天，我来尝试使用PaddleFluid做图像分类工作。这里需要说明的是，PaddleFluid更新的频率太高了，我这里的代码是在0.13.0的基础上写的，很多更方便的API，我在官方github上的branch看到了，但是暂时无法使用，不过我这里也会告诉大家这些新的写法。Fluid文档太少，需要看代码研究python接口。
模型介绍
resnet 搞过图像的应该都知道，kaiminghe的resnet，
Deep Residual Learning for Image Recognition (15年年底的文章，竟然有9496个citations)，就是那个最开始搞过1000多层的网络的，原理不说了，有好多好多的文章有介绍，随便一搜就好啦。
这里，不需要把resnet的每一层用fluid都写出来，PaddlePaddle的repo里面有这块的工作，可供直接复用。
resnet.py
Dog vs Cat
Kaggle网站上找到Dog vs Cat 数据集，
Dogs vs. Cats, 安装好kaggle-api 后kaggle competitions download -c dogs-vs-cats 即可下载数据集，后面实验我在训练集中用80%做训练数据，20%做验证集。
Image Reader
def default_mapper(sample):
img, label = sample
img = image.simple_transform(
img, 256, 224, True, mean=[103.94, 116.78, 123.68])
return img.flatten().astype('float32'), label
def dataset_reader(data_dir, train_val_ratio=0.8):
img_list = []
img2label = dict()
label2id = dict()
sub_dirs = [i for i in os.listdir(data_dir) if os.path.isdir(i)]
for index, sub_dir in enumerate(sub_dirs):
label2id[sub_dir] = index
sub_files = []
for root, dir, files in os.walk(os.path.join(data_dir, sub_dir)):
sub_files = [os.path.join(root, file) for file in files if file.split(".")[-1] in ["jpg, jpeg"]]
img_list += sub_files
for file in sub_files:
img2label[file] =sub_dir
random.shuffle(img_list)
train_len = int(train_val_ratio*len(img_list))
train_img_list = img_list[:train_len]
val_img_list = img_list[train_len:]
def train_reader():
for idx, imgfile in enumerate(train_img_list):
try:
data = image.load_image(imgfile)
label = [label2id[img2label[imgfile]], ]
yield [data, label]
except Exception as e:
print "error infor: {0}".format(e.message)
continue
def test_reader():
for idx, imgfile in enumerate(val_img_list):
try:
data = image.load_image(imgfile)
label = [label2id[img2label[imgfile]], ]
yield [data, label]
except Exception as e:
print "error infor: {0}".format(e.message)
continue
return paddle.reader.map_readers(default_mapper, train_reader), paddle.reader.map_readers(default_mapper, test_reader)
data_reader函数主要有两个部分：
遍历所有图像;
读取图像，生成train，test的生成器;
模型构建
def conv_bn_layer(input, ch_out, filter_size, stride, padding, act='relu'):
conv1 = fluid.layers.conv2d(
input=input,
filter_size=filter_size,
num_filters=ch_out,
stride=stride,
padding=padding,
act=None,
bias_attr=False)
return fluid.layers.batch_norm(input=conv1, act=act)
def shortcut(input, ch_out, stride):
ch_in = input.shape[1] # if args.data_format == 'NCHW' else input.shape[-1]
if ch_in != ch_out:
return conv_bn_layer(input, ch_out, 1, stride, 0, None)
else:
return input
def basicblock(input, ch_out, stride):
short = shortcut(input, ch_out, stride)
conv1 = conv_bn_layer(input, ch_out, 3, stride, 1)
conv2 = conv_bn_layer(conv1, ch_out, 3, 1, 1, act=None)
return fluid.layers.elementwise_add(x=short, y=conv2, act='relu')
def bottleneck(input, ch_out, stride):
short = shortcut(input, ch_out * 4, stride)
conv1 = conv_bn_layer(input, ch_out, 1, stride, 0)
conv2 = conv_bn_layer(conv1, ch_out, 3, 1, 1)
conv3 = conv_bn_layer(conv2, ch_out * 4, 1, 1, 0, act=None)
return fluid.layers.elementwise_add(x=short, y=conv3, act='relu')
def layer_warp(block_func, input, ch_out, count, stride):
res_out = block_func(input, ch_out, stride)
for i in range(1, count):
res_out = block_func(res_out, ch_out, 1)
return res_out
def resnet(input, class_dim, depth=18, data_format='NCHW'):
cfg = {
18: ([2, 2, 2, 1], basicblock),
34: ([3, 4, 6, 3], basicblock),
50: ([3, 4, 6, 3], bottleneck),
101: ([3, 4, 23, 3], bottleneck),
152: ([3, 8, 36, 3], bottleneck)
}
stages, block_func = cfg[depth]
conv1 = conv_bn_layer(input, ch_out=64, filter_size=7, stride=2, padding=3)
pool1 = fluid.layers.pool2d(
input=conv1, pool_type='avg', pool_size=3, pool_stride=2)
res1 = layer_warp(block_func, pool1, 64, stages[0], 1)
res2 = layer_warp(block_func, res1, 128, stages[1], 2)
res3 = layer_warp(block_func, res2, 256, stages[2], 2)
res4 = layer_warp(block_func, res3, 512, stages[3], 2)
pool2 = fluid.layers.pool2d(
input=res4,
pool_size=7,
pool_type='avg',
pool_stride=1,
global_pooling=True)
out = fluid.layers.fc(input=pool2, size=class_dim, act='softmax')
return out
resnet()配置不同层数的resnet网络，如resnet50，resnet34, resnet101等，这里主要是fluid的api，主要是和模型结构相关的，一般来说，经典的模型都会有重现，想使用的同学google一下会有相应的实现，当然也要理解下怎么做的，这里我就不深究了，对比着论文应该不难。
训练
def train(args):
# logger = LogWriter(args.logdir, sync_cycle=10000)
model = resnet
class_dim = args.class_dim
if args.data_format == 'NCHW':
dshape = [3, 224, 224]
else:
dshape = [224, 224, 3]
if not args.data_path:
raise Exception(
"Must specify --data_path when training with imagenet")
train_reader, test_reader = dataset_reader(args.data_path)
print(train_reader)
def train_network():
input = fluid.layers.data(name='image', shape=dshape, dtype='float32')
predict = model(input, class_dim)
label = fluid.layers.data(name='label', shape=[1], dtype='int64')
cost = fluid.layers.cross_entropy(input=predict, label=label)
avg_cost = fluid.layers.mean(x=cost)
batch_acc = fluid.layers.accuracy(input=predict, label=label)
return [avg_cost, batch_acc]
optimizer = fluid.optimizer.Momentum(learning_rate=0.01, momentum=0.9)
batched_train_reader = paddle.batch(
paddle.reader.shuffle(
train_reader, buf_size=5120),
batch_size=args.batch_size
)
batched_test_reader = paddle.batch(
test_reader, batch_size=args.batch_size)
def event_handler(event):
if isinstance(event, fluid.EndStepEvent):
print('Pass:{0},Step: {1},Metric: {2}'.format(event.epoch, event.step, event.metrics))
if isinstance(event, fluid.EndEpochEvent):
# save model to dir
#trainer.save_params(".")
avg_cost, acc = trainer.test(reader=batched_test_reader, feed_order=["image", "label"])
print('Pass:{0},val avg_cost: {1}, acc: {2}'.format(event.epoch, avg_cost, acc))
trainer.save_params("./ckpt")
# write the loss, acc to visualdl file
pass
# place = fluid.CUDAPlace(0) if args.use_cuda else fluid.CPUPlace()
place = fluid.CUDAPlace(0)
trainer = fluid.Trainer(
train_func=train_network, optimizer=optimizer, place=place)
print("Begin to Train")
trainer.train(
reader=batched_train_reader,
num_epochs=args.pass_num,
event_handler=event_handler,
feed_order=['image', 'label'])
train()主要包括:
构建模型，主要是resnet的部分，构建各种不同的layer,直接使用上节的模型构造即可;
构建训练相关的部分，配置输入输出(input，label),构建cost，acc这类op;
将train_reader, test_reader包装为batch_reader;
配置设备信息，新建Trainer开始训练;
epoch结束后保存模型的部分还是使用v2的风格，github中Fluid已经支持CheckpointConfig来完成相应的配置，传给Trainer，但是我这边应该从pip安装的是0.13.0的版本，我进系统的文件看了下这部分更改没有更新，所以就先使用v2的风格，save_params来保存模型参数，个人从技术角度来说更偏爱CheckpointConfig这种config的模式。
训练日志
训练过程中发现一点问题：GPU占用率跳动比较频繁， 占用率经常跳到0，怀疑是等待问题，看代码部分发现paddle.reader.map_readers(default_mapper, train_reader)没有配置多个线程， 应该是由于单个线程在读image，包括预处理的部分时间过长，造成了gpu计算时间的等待, 修改为paddle.reader.xmap_readers(default_mapper, train_reader, cpu_count(), 51200)之后，运行快了很多，不过还是有比较明显的GPU占用率跳的比较明显，看了下源码，读取数据的部分是python实现的，并不是很高效，现在只有一张卡，还好，要是多张卡，等待会更明显，这部分应该有一个更好的替代方案，可以从底层cpp来实现相应的读取逻辑，效率会很高。
def xmap_readers(mapper, reader, process_num, buffer_size, order=False):
end = XmapEndSignal()
# define a worker to read samples from reader to in_queue
def read_worker(reader, in_queue):
for i in reader():
in_queue.put(i)
in_queue.put(end)
# define a worker to read samples from reader to in_queue with order flag
def order_read_worker(reader, in_queue):
in_order = 0
for i in reader():
in_queue.put((in_order, i))
in_order += 1
in_queue.put(end)
# define a worker to handle samples from in_queue by mapper
# and put mapped samples into out_queue
def handle_worker(in_queue, out_queue, mapper):
sample = in_queue.get()
while not isinstance(sample, XmapEndSignal):
r = mapper(sample)
out_queue.put(r)
sample = in_queue.get()
in_queue.put(end)
out_queue.put(end)
# define a worker to handle samples from in_queue by mapper
# and put mapped samples into out_queue by order
def order_handle_worker(in_queue, out_queue, mapper, out_order):
ins = in_queue.get()
while not isinstance(ins, XmapEndSignal):
order, sample = ins
r = mapper(sample)
while order != out_order[0]:
pass
out_queue.put(r)
out_order[0] += 1
ins = in_queue.get()
in_queue.put(end)
out_queue.put(end)
def xreader():
in_queue = Queue(buffer_size)
out_queue = Queue(buffer_size)
out_order = [0]
# start a read worker in a thread
target = order_read_worker if order else read_worker
t = Thread(target=target, args=(reader, in_queue))
t.daemon = True
t.start()
# start several handle_workers
target = order_handle_worker if order else handle_worker
args = (in_queue, out_queue, mapper, out_order) if order else (
in_queue, out_queue, mapper)
workers = []
for i in xrange(process_num):
worker = Thread(target=target, args=args)
worker.daemon = True
workers.append(worker)
for w in workers:
w.start()
sample = out_queue.get()
while not isinstance(sample, XmapEndSignal):
yield sample
sample = out_queue.get()
finish = 1
while finish < process_num:
sample = out_queue.get()
if isinstance(sample, XmapEndSignal):
finish += 1
else:
yield sample
return xreader
Image Augmentation
前面，我简单地跑起了流程，没有做基本的处理，比如Image Augmentation，如果做了Image Augmentation, 效果应该会更好一些，这里测试一下Image Augmentation。
读下上面的代码， Image Augmentation的部分可以在default_maper的部分实现，这里可以尝试下：
DATA_DIM=224
img_mean = np.array([0.485, 0.456, 0.406]).reshape((3, 1, 1))
img_std = np.array([0.229, 0.224, 0.225]).reshape((3, 1, 1))
def resize_short(img, target_size):
percent = float(target_size) / min(img.size[0], img.size[1])
resized_color = int(round(img.size[0] * percent))
resized_color = int(round(img.size[1] * percent))
img = img.resize((resized_color, resized_color), Image.LANCZOS)
return img
def crop_image(img, target_size, center):
color, color = img.size
size = target_size
if center == True:
w_start = (color - size) / 2
h_start = (color - size) / 2
else:
w_start = random.randint(0, color - size)
h_start = random.randint(0, color - size)
w_end = w_start + size
h_end = h_start + size
img = img.crop((w_start, h_start, w_end, h_end))
return img
def random_crop(img, size, scale=[0.08, 1.0], ratio=[3. / 4., 4. / 3.]):
aspect_ratio = math.sqrt(random.uniform(*ratio))
w = 1. * aspect_ratio
h = 1. / aspect_ratio
bound = min((float(img.size[0]) / img.size[1]) / (w**2),
(float(img.size[1]) / img.size[0]) / (h**2))
scale_max = min(scale[1], bound)
scale_min = min(scale[0], bound)
target_area = img.size[0] * img.size[1] * random.uniform(scale_min,
scale_max)
target_size = math.sqrt(target_area)
w = int(target_size * w)
h = int(target_size * h)
i = random.randint(0, img.size[0] - w)
j = random.randint(0, img.size[1] - h)
img = img.crop((i, j, i + w, j + h))
img = img.resize((size, size), Image.LANCZOS)
return img
def rotate_image(img):
angle = random.randint(-10, 10)
img = img.rotate(angle)
return img
def distort_color(img):
def random_brightness(img, lower=0.5, upper=1.5):
e = random.uniform(lower, upper)
return ImageEnhance.Brightness(img).enhance(e)
def random_contrast(img, lower=0.5, upper=1.5):
e = random.uniform(lower, upper)
return ImageEnhance.Contrast(img).enhance(e)
def random_color(img, lower=0.5, upper=1.5):
e = random.uniform(lower, upper)
return ImageEnhance.Color(img).enhance(e)
ops = [random_brightness, random_contrast, random_color]
random.shuffle(ops)
img = ops[0](img)
img = ops[1](img)
img = ops[2](img)
return img
def process_image(sample, mode, color_jitter, rotate):
img_path = sample[0]
img = Image.open(img_path)
#img = sample[0]
if mode == 'train':
if rotate: img = rotate_image(img)
img = random_crop(img, DATA_DIM)
else:
img = resize_short(img, target_size=256)
img = crop_image(img, target_size=DATA_DIM, center=True)
if mode == 'train':
if color_jitter:
img = distort_color(img)
if random.randint(0, 1) == 1:
img = img.transpose(Image.FLIP_LEFT_RIGHT)
if img.mode != 'RGB':
img = img.convert('RGB')
img = np.array(img).astype('float32').transpose((2, 0, 1)) / 255
img -= img_mean
img /= img_std
if mode == 'train' or mode == 'val':
return img, sample[1]
elif mode == 'test':
return [img]
然后修改mapper的部分：
train_mapper = functools.partial(process_image, mode="train", color_jitter=False, rotate=False)
test_mapper = functools.partial(process_image, mode="test")
return paddle.reader.xmap_readers(train_mapper, train_reader, cpu_count(), 51200), paddle.reader.xmap_readers(test_mapper, test_reader, cpu_count(), 5120)
这里可以对比一下Image Augmentation前后的、在验证集上的结果：
很明显，在完成Image Aug之后，结果有了进一步提升。
所有的源码都更新在paddle-101， 这里只做基本的demo，成绩这块未做进一步工作，大家可以尝试用Fluid刷下榜看看， 有兴趣的小伙伴可以玩一玩。
