Hola Habr! Les presento a ustedes la traducción del artículo "Implementación de la agrupación de ROI en TensorFlow + Keras" de Jaime Sevilla.

Actualmente estoy tomando un curso de aprendizaje automático. En el bloque de capacitación "Visión por computadora" era necesario estudiar el RoI Pooling de capas. El siguiente artículo me pareció interesante y, por lo tanto, decidí compartir la traducción con la comunidad.

En esta publicación, explicaremos el concepto básico y el uso general de la agrupación de RoI ( Región de interés ) y ofreceremos una implementación utilizando capas TensorFlow Keras.

El público objetivo de esta publicación son personas que están familiarizadas con la teoría básica de las redes neuronales (convolucionales) (CNN) y son capaces de construir y ejecutar modelos simples usando Keras .

Si solo está aquí por el código, marque aquí y no se olvide de dar me gusta y compartir el artículo.

Comprender la agrupación de ROI

La agrupación de ROI fue propuesta por Ross Girshik en el artículo Fast R-CNN como parte de su canal de reconocimiento de objetos.

En un caso de uso general para la agrupación de ROI , tenemos un objeto similar a una imagen y varias regiones de interés ( RoI ) especificadas a través de cuadros delimitadores. Queremos crear incrustaciones (incrustaciones - mapear una entidad arbitraria (una pieza de una imagen) a un determinado vector) de cada ROI.

Por ejemplo, en una configuración R-CNN, tenemos una imagen y un motor de resaltado de región candidato que crea cuadros delimitadores para partes potencialmente interesantes de la imagen. Ahora queremos crear una incrustación para cada pieza sugerida de la imagen.

Simplemente recortar cada área sugerida no funcionará porque queremos superponer las incrustaciones resultantes una encima de la otra, ¡y las áreas sugeridas no necesariamente tienen la misma forma!

, . ?

- (pooling).

max pooling, ( ) , , .

maxpool

, – : .

. RoI ?

ROI Pooling , pooling.

, ROI Pooling.

RoI Pooling.

RoI Pooling — . , RoI, . , .

-, ( RoI Pooling ), ( ), (end-to-end) (single-pass) .

Fast R-CNN, RoI Pooling,

, R-CNN , (RoI). RoI Pooling CNN . .

-, -, RoI Pooling (visual attention).

Attentional Network , ROI Pooling, Hara ..

Attentional Network , Hara attention, ROI ROI Pooling. (t = 1), ROI Pooling (Fully Connected). Glimpse () (t = 2) , ROI Pooling. .

.

, , ROI.

:

• (batch) . , . (batch_size, img_width, img_height, n_channels), batch_size- , img_width — , img_height — , n_channels — .
• (batch) ROI. , - . 4 , (batch_size, n_rois, 4), batch_size — ROI, n_rois — ROI.

:

• , ROI. (batch_size, n_rois, pooled_width, pooled_height, n_channels). batch_size- , n_rois — ROI, pooled_width — , pooled_height— , n_channels — .

Keras

Keras Layer.

tf.keras init, build call . , build , , . compute_output_shape.

, .

def __init__(self, pooled_height, pooled_width, **kwargs):
    self.pooled_height = pooled_height
    self.pooled_width = pooled_width
    super(ROIPoolingLayer, self).__init__(**kwargs)

. , . .

def compute_output_shape(self, input_shape):
    """ Returns the shape of the ROI Layer output
    """
    feature_map_shape, rois_shape = input_shape
    assert feature_map_shape[0] == rois_shape[0]
    batch_size = feature_map_shape[0]
    n_rois = rois_shape[1]
    n_channels = feature_map_shape[3]
    return (batch_size, n_rois, self.pooled_height, 
            self.pooled_width, n_channels)

compute_output_shape — , , .

(call). — , . , ROI Pooling, .

, , ROI .

.

@staticmethod
def _pool_roi(feature_map, roi, pooled_height, pooled_width):
  """ Applies ROI Pooling to a single image and a single ROI
  """# Compute the region of interest        
  feature_map_height = int(feature_map.shape[0])
  feature_map_width  = int(feature_map.shape[1])

  h_start = tf.cast(feature_map_height * roi[0], 'int32')
  w_start = tf.cast(feature_map_width  * roi[1], 'int32')
  h_end   = tf.cast(feature_map_height * roi[2], 'int32')
  w_end   = tf.cast(feature_map_width  * roi[3], 'int32')

  region = feature_map[h_start:h_end, w_start:w_end, :]
...

, .

, ROI , 0 1. , ROI 4- , (x_min, y_min, x_max, y_max ).

ROI , , , : , , ROI Pooling, , , ROI.

, TensorFlow.

...
# Divide the region into non overlapping areas
region_height = h_end - h_start
region_width  = w_end - w_start
h_step = tf.cast(region_height / pooled_height, 'int32')
w_step = tf.cast(region_width  / pooled_width , 'int32')

areas = [[(
           i*h_step, 
           j*w_step, 
           (i+1)*h_step if i+1 < pooled_height else region_height, 
           (j+1)*w_step if j+1 < pooled_width else region_width
          ) 
          for j in range(pooled_width)] 
         for i in range(pooled_height)]
...

ROI, .

2D , , , .

, , , , ROI (region_height // pooled_height, region_width // pooled_width), ROI, .

2D , .

...
# Take the maximum of each area and stack the result
def pool_area(x): 
  return tf.math.reduce_max(region[x[0]:x[2],x[1]:x[3],:], axis=[0,1])

pooled_features = tf.stack([[pool_area(x) for x in row] for row in areas])
return pooled_features

. pool_area, , , , , .

pool_area , , list comprehension .

(pooled_height, pooled_width, n_channels), RoI .

— RoI . tf.map_fn (n_rois, pooled_height, pooled_width, n_channels).

@staticmethod
def _pool_rois(feature_map, rois, pooled_height, pooled_width):
  """ Applies ROI pooling for a single image and varios ROIs
  """
  def curried_pool_roi(roi): 
    return ROIPoolingLayer._pool_roi(feature_map, roi, 
                                     pooled_height, pooled_width)

  pooled_areas = tf.map_fn(curried_pool_roi, rois, dtype=tf.float32)
  return pooled_areas

, . tf.map_fn (, x), , .

def call(self, x):
  """ Maps the input tensor of the ROI layer to its output
  """
  def curried_pool_rois(x): 
    return ROIPoolingLayer._pool_rois(x[0], x[1], 
                                      self.pooled_height, 
                                      self.pooled_width)

  pooled_areas = tf.map_fn(curried_pool_rois, x, dtype=tf.float32)
  return pooled_areas

, dtype tf.map_fn , . , , , Tensorflow.

:

import tensorflow as tf
from tensorflow.keras.layers import Layer

class ROIPoolingLayer(Layer):
    """ Implements Region Of Interest Max Pooling 
        for channel-first images and relative bounding box coordinates

        # Constructor parameters
            pooled_height, pooled_width (int) -- 
              specify height and width of layer outputs

        Shape of inputs
            [(batch_size, pooled_height, pooled_width, n_channels),
             (batch_size, num_rois, 4)]

        Shape of output
            (batch_size, num_rois, pooled_height, pooled_width, n_channels)

    """
    def __init__(self, pooled_height, pooled_width, **kwargs):
        self.pooled_height = pooled_height
        self.pooled_width = pooled_width

        super(ROIPoolingLayer, self).__init__(**kwargs)

    def compute_output_shape(self, input_shape):
        """ Returns the shape of the ROI Layer output
        """
        feature_map_shape, rois_shape = input_shape
        assert feature_map_shape[0] == rois_shape[0]
        batch_size = feature_map_shape[0]
        n_rois = rois_shape[1]
        n_channels = feature_map_shape[3]
        return (batch_size, n_rois, self.pooled_height, 
                self.pooled_width, n_channels)

    def call(self, x):
        """ Maps the input tensor of the ROI layer to its output

            # Parameters
                x[0] -- Convolutional feature map tensor,
                        shape (batch_size, pooled_height, pooled_width, n_channels)
                x[1] -- Tensor of region of interests from candidate bounding boxes,
                        shape (batch_size, num_rois, 4)
                        Each region of interest is defined by four relative 
                        coordinates (x_min, y_min, x_max, y_max) between 0 and 1
            # Output
                pooled_areas -- Tensor with the pooled region of interest, shape
                    (batch_size, num_rois, pooled_height, pooled_width, n_channels)
        """
        def curried_pool_rois(x): 
          return ROIPoolingLayer._pool_rois(x[0], x[1], 
                                            self.pooled_height, 
                                            self.pooled_width)

        pooled_areas = tf.map_fn(curried_pool_rois, x, dtype=tf.float32)

        return pooled_areas

    @staticmethod
    def _pool_rois(feature_map, rois, pooled_height, pooled_width):
        """ Applies ROI pooling for a single image and varios ROIs
        """
        def curried_pool_roi(roi): 
          return ROIPoolingLayer._pool_roi(feature_map, roi, 
                                           pooled_height, pooled_width)

        pooled_areas = tf.map_fn(curried_pool_roi, rois, dtype=tf.float32)
        return pooled_areas

    @staticmethod
    def _pool_roi(feature_map, roi, pooled_height, pooled_width):
        """ Applies ROI pooling to a single image and a single region of interest
        """

        # Compute the region of interest        
        feature_map_height = int(feature_map.shape[0])
        feature_map_width  = int(feature_map.shape[1])

        h_start = tf.cast(feature_map_height * roi[0], 'int32')
        w_start = tf.cast(feature_map_width  * roi[1], 'int32')
        h_end   = tf.cast(feature_map_height * roi[2], 'int32')
        w_end   = tf.cast(feature_map_width  * roi[3], 'int32')

        region = feature_map[h_start:h_end, w_start:w_end, :]

        # Divide the region into non overlapping areas
        region_height = h_end - h_start
        region_width  = w_end - w_start
        h_step = tf.cast( region_height / pooled_height, 'int32')
        w_step = tf.cast( region_width  / pooled_width , 'int32')

        areas = [[(
                    i*h_step, 
                    j*w_step, 
                    (i+1)*h_step if i+1 < pooled_height else region_height, 
                    (j+1)*w_step if j+1 < pooled_width else region_width
                   ) 
                   for j in range(pooled_width)] 
                  for i in range(pooled_height)]

        # take the maximum of each area and stack the result
        def pool_area(x): 
          return tf.math.reduce_max(region[x[0]:x[2], x[1]:x[3], :], axis=[0,1])

        pooled_features = tf.stack([[pool_area(x) for x in row] for row in areas])
        return pooled_features

! , 1- 100x200, 2 RoI, 7x3. , 4 . — 1, 50 (-1, -3).

import numpy as np# Define parameters
batch_size = 1
img_height = 200
img_width = 100
n_channels = 1
n_rois = 2
pooled_height = 3
pooled_width = 7# Create feature map input
feature_maps_shape = (batch_size, img_height, img_width, n_channels)
feature_maps_tf = tf.placeholder(tf.float32, shape=feature_maps_shape)
feature_maps_np = np.ones(feature_maps_tf.shape, dtype='float32')
feature_maps_np[0, img_height-1, img_width-3, 0] = 50
print(f"feature_maps_np.shape = {feature_maps_np.shape}")# Create batch size
roiss_tf = tf.placeholder(tf.float32, shape=(batch_size, n_rois, 4))
roiss_np = np.asarray([[[0.5,0.2,0.7,0.4], [0.0,0.0,1.0,1.0]]], dtype='float32')
print(f"roiss_np.shape = {roiss_np.shape}")# Create layer
roi_layer = ROIPoolingLayer(pooled_height, pooled_width)
pooled_features = roi_layer([feature_maps_tf, roiss_tf])
print(f"output shape of layer call = {pooled_features.shape}")# Run tensorflow session
with tf.Session() as session:
    result = session.run(pooled_features, 
                         feed_dict={feature_maps_tf:feature_maps_np,  
                                    roiss_tf:roiss_np})

print(f"result.shape = {result.shape}")
print(f"first  roi embedding=\n{result[0,0,:,:,0]}")
print(f"second roi embedding=\n{result[0,1,:,:,0]}")

, TensorFlow, .

:

feature_maps_np.shape = (1, 200, 100, 1)
roiss_np.shape = (1, 2, 4)
output shape of layer call = (1, 2, 3, 7, 1)
result.shape = (1, 2, 3, 7, 1)
first  roi embedding=
[[1. 1. 1. 1. 1. 1. 1.]
 [1. 1. 1. 1. 1. 1. 1.]
 [1. 1. 1. 1. 1. 1. 1.]]
second roi embedding=
[[ 1.  1.  1.  1.  1.  1.  1.]
 [ 1.  1.  1.  1.  1.  1.  1.]
 [ 1.  1.  1.  1.  1.  1. 50.]]

, . — 1, , 50.

, !

, !

, ROI Pooling (attention). , , Keras , ROI Pooling .

, , , !

Ari Brill, Tjark Miener Bryan Kim .

• Ross Girshick. Fast R-CNN. Proceedings of the IEEE International Conference on Computer Vision. 2015.
• Kota Hara, Ming-Yu Liu, Oncel Tuzel, Amir-massoud Farahmand. Attentional Network for Visual Object Detection. 2017.