deepboof.forward.SpatialPooling Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of main
Trainer Agnostic Deep Learning
There is a newer version: 0.5.3
/*
 * Copyright (c) 2016, Peter Abeles. All Rights Reserved.
 *
 * This file is part of DeepBoof
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *   http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package deepboof.forward;

import deepboof.Function;
import deepboof.Tensor;

import java.util.List;

/**
 * Spatial pooling down samples the input spatial tensors by finding a representative value inside
 * each pooling region.  The intent is to reduce the number of variables while maintaining much of the original
 * information.  Pooling is specified by the region's size (poolWidth, poolHeight) and the sampling
 * period (periodX, periodY), and padding parameters.
 *
 * Processing Steps:

 * 
 *   Apply spatial padding
 *   Apply spatial pooling to padded image
 * 
 *
 * Notes
 * 
 *   The first region has it's lower extent at the spatial region's lower extent.
 *   Sampling is done in a row-major ordering, e.g. columns then rows
 *   If a region extends outside the image plus padding it's ignore
 * 
 *
 * @see SpatialPadding2D
 * @see SpatialMaxPooling
 * @see SpatialAveragePooling
 *
 * @author Peter Abeles
 */
public interface SpatialPooling extends Function {

	/**
	 * Processes a spatial tensor.
	 *
	 * 	 * N  = number of mini-batch images
	 * C  = number of channels in each image
	 * H  = height of input image
	 * W  = width of input image
	 * Hp = height of input image + padding
	 * Wp = width of input image + padding
	 *
	 * Shape of output spacial tensor:
	 *
	 * H' = 1 + (Hp - poolHeight) / periodY
	 * W' = 1 + (Wp - poolWidth ) / periodX
	 * 
	 *
	 * @param input Input spacial tensor = (N, C, H, W)
	 * @param output Output spatial tensor = (N, C, H', W')
	 */
	@Override
	void forward(T input, T output);

	/**
	 * Can skip.  No parameters required.
	 *
	 * @param parameters No parameters required
	 */
	@Override
	void setParameters(List parameters);

	/**
	 * Returns pooling configuration
	 * @return configuration
	 */
	ConfigSpatial getConfiguration();
}