UNPKG

@tensorflow-models/coco-ssd

Version:

Object detection model (coco-ssd) in TensorFlow.js

381 lines 20.4 kB
"use strict"; /** * @license * Copyright 2018 Google Inc. All Rights Reserved. * 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. * ============================================================================= */ Object.defineProperty(exports, "__esModule", { value: true }); var environment_1 = require("../environment"); var tensor_util_env_1 = require("../tensor_util_env"); var util = require("../util"); var array_ops_1 = require("./array_ops"); var conv_util = require("./conv_util"); var operation_1 = require("./operation"); /** * Computes the 2D max pooling of an image. * * @param x The input tensor, of rank 4 or rank 3 of shape * `[batch, height, width, inChannels]`. If rank 3, batch of 1 is assumed. * @param filterSize The filter size: `[filterHeight, filterWidth]`. If * `filterSize` is a single number, then `filterHeight == filterWidth`. * @param strides The strides of the pooling: `[strideHeight, strideWidth]`. If * `strides` is a single number, then `strideHeight == strideWidth`. * @param dilations The dilation rates: `[dilationHeight, dilationWidth]` * in which we sample input values across the height and width dimensions * in dilated pooling. Defaults to `[1, 1]`. If `dilations` is a single * number, then `dilationHeight == dilationWidth`. If it is greater than * 1, then all values of `strides` must be 1. * @param pad The type of padding algorithm. * - `same` and stride 1: output will be of same size as input, * regardless of filter size. * - `valid`: output will be smaller than input if filter is larger * than 1x1. * - For more info, see this guide: * [https://www.tensorflow.org/api_guides/python/nn#Convolution]( * https://www.tensorflow.org/api_guides/python/nn#Convolution) * @param dimRoundingMode The rounding mode used when computing output * dimensions if pad is a number. If none is provided, it will not round * and error if the output is of fractional size. */ function maxPoolImpl_(x, filterSize, strides, dilations, pad, dimRoundingMode) { var $x = tensor_util_env_1.convertToTensor(x, 'x', 'maxPool'); var x4D = $x; var reshapedTo4D = false; if ($x.rank === 3) { reshapedTo4D = true; x4D = $x.as4D(1, $x.shape[0], $x.shape[1], $x.shape[2]); } if (dilations == null) { dilations = [1, 1]; } util.assert(x4D.rank === 4, "Error in maxPool: input must be rank 4 but got rank " + x4D.rank + "."); util.assert(conv_util.eitherStridesOrDilationsAreOne(strides, dilations), 'Error in maxPool: Either strides or dilations must be 1. ' + ("Got strides " + strides + " and dilations '" + dilations + "'")); if (dimRoundingMode != null) { util.assert(util.isInt(pad), "Error in maxPool: pad must be an integer when using, " + ("dimRoundingMode " + dimRoundingMode + " but got pad " + pad + ".")); } var convInfo = conv_util.computePool2DInfo(x4D.shape, filterSize, strides, dilations, pad, dimRoundingMode); var grad = function (dy, saved) { var y4D = saved[0]; return { x: function () { return maxPoolBackprop(dy, x4D, y4D, filterSize, strides, dilations, pad); } }; }; var res = environment_1.ENV.engine.runKernel(function (backend, save) { return save(backend.maxPool(x4D, convInfo)); }, { x: x4D }, grad); if (reshapedTo4D) { return res.as3D(res.shape[1], res.shape[2], res.shape[3]); } return res; } /** * Computes the 2D max pooling of an image. * * @param x The input tensor, of rank 4 or rank 3 of shape * `[batch, height, width, inChannels]`. If rank 3, batch of 1 is assumed. * @param filterSize The filter size: `[filterHeight, filterWidth]`. If * `filterSize` is a single number, then `filterHeight == filterWidth`. * @param strides The strides of the pooling: `[strideHeight, strideWidth]`. If * `strides` is a single number, then `strideHeight == strideWidth`. * @param pad The type of padding algorithm. * - `same` and stride 1: output will be of same size as input, * regardless of filter size. * - `valid`: output will be smaller than input if filter is larger * than 1x1. * - For more info, see this guide: * [https://www.tensorflow.org/api_guides/python/nn#Convolution]( * https://www.tensorflow.org/api_guides/python/nn#Convolution) * @param dimRoundingMode The rounding mode used when computing output * dimensions if pad is a number. If none is provided, it will not round * and error if the output is of fractional size. */ /** @doc {heading: 'Operations', subheading: 'Convolution'} */ function maxPool_(x, filterSize, strides, pad, dimRoundingMode) { return maxPoolImpl_(x, filterSize, strides, 1, pad, dimRoundingMode); } /** * Computes the 2D average pooling of an image. * * @param x The input tensor, of rank 4 or rank 3 of shape * `[batch, height, width, inChannels]`. If rank 3, batch of 1 is assumed. * @param filterSize The filter size: `[filterHeight, filterWidth]`. If * `filterSize` is a single number, then `filterHeight == filterWidth`. * @param strides The strides of the pooling: `[strideHeight, strideWidth]`. If * `strides` is a single number, then `strideHeight == strideWidth`. * @param dilations The dilation rates: `[dilationHeight, dilationWidth]` * in which we sample input values across the height and width dimensions * in dilated pooling. Defaults to `[1, 1]`. If `dilations` is a single * number, then `dilationHeight == dilationWidth`. If it is greater than * 1, then all values of `strides` must be 1. * @param pad The type of padding algorithm: * - `same` and stride 1: output will be of same size as input, * regardless of filter size. * - `valid`: output will be smaller than input if filter is larger * than 1x1. * - For more info, see this guide: * [https://www.tensorflow.org/api_guides/python/nn#Convolution]( * https://www.tensorflow.org/api_guides/python/nn#Convolution) * @param dimRoundingMode The rounding mode used when computing output * dimensions if pad is a number. If none is provided, it will not round * and error if the output is of fractional size. */ function avgPoolImpl_(x, filterSize, strides, dilations, pad, dimRoundingMode) { var $x = tensor_util_env_1.convertToTensor(x, 'x', 'avgPool', 'float32'); if (dilations == null) { dilations = [1, 1]; } util.assert(conv_util.eitherStridesOrDilationsAreOne(strides, dilations), 'Error in avgPool: Either strides or dilations must be 1. ' + ("Got strides " + strides + " and dilations '" + dilations + "'")); var x4D = $x; var reshapedTo4D = false; if ($x.rank === 3) { reshapedTo4D = true; x4D = $x.as4D(1, $x.shape[0], $x.shape[1], $x.shape[2]); } util.assert(x4D.rank === 4, "Error in avgPool: x must be rank 4 but got rank " + x4D.rank + "."); if (dimRoundingMode != null) { util.assert(util.isInt(pad), "Error in avgPool: pad must be an integer when using, " + ("dimRoundingMode " + dimRoundingMode + " but got pad " + pad + ".")); } var convInfo = conv_util.computePool2DInfo(x4D.shape, filterSize, strides, dilations, pad, dimRoundingMode); var grad = function (dy) { return { x: function () { return avgPoolBackprop(dy, x4D, filterSize, strides, dilations, pad); } }; }; var res = environment_1.ENV.engine.runKernel(function (backend) { return backend.avgPool(x4D, convInfo); }, { x: x4D }, grad); res = res.cast($x.dtype); if (reshapedTo4D) { return res.as3D(res.shape[1], res.shape[2], res.shape[3]); } return res; } /** * Computes the 2D average pooling of an image. * * @param x The input tensor, of rank 4 or rank 3 of shape * `[batch, height, width, inChannels]`. If rank 3, batch of 1 is assumed. * @param filterSize The filter size: `[filterHeight, filterWidth]`. If * `filterSize` is a single number, then `filterHeight == filterWidth`. * @param strides The strides of the pooling: `[strideHeight, strideWidth]`. If * `strides` is a single number, then `strideHeight == strideWidth`. * @param pad The type of padding algorithm: * - `same` and stride 1: output will be of same size as input, * regardless of filter size. * - `valid`: output will be smaller than input if filter is larger * than 1x1. * - For more info, see this guide: * [https://www.tensorflow.org/api_guides/python/nn#Convolution]( * https://www.tensorflow.org/api_guides/python/nn#Convolution) * @param dimRoundingMode The rounding mode used when computing output * dimensions if pad is a number. If none is provided, it will not round * and error if the output is of fractional size. */ /** @doc {heading: 'Operations', subheading: 'Convolution'} */ function avgPool_(x, filterSize, strides, pad, dimRoundingMode) { return avgPoolImpl_(x, filterSize, strides, 1, pad, dimRoundingMode); } /** * Performs an N-D pooling operation * * @param input The input tensor, of rank 4 or rank 3 of shape * `[batch, height, width, inChannels]`. If rank 3, batch of 1 is assumed. * @param windowShape The filter size: `[filterHeight, filterWidth]`. If * `filterSize` is a single number, then `filterHeight == filterWidth`. * @param poolingType The type of pooling, either 'max' or 'avg'. * @param pad The type of padding algorithm: * - `same` and stride 1: output will be of same size as input, * regardless of filter size. * - `valid`: output will be smaller than input if filter is larger * than 1x1. * - For more info, see this guide: * [https://www.tensorflow.org/api_guides/python/nn#Convolution]( * https://www.tensorflow.org/api_guides/python/nn#Convolution) * @param dilations The dilation rates: `[dilationHeight, dilationWidth]` * in which we sample input values across the height and width dimensions * in dilated pooling. Defaults to `[1, 1]`. If `dilationRate` is a single * number, then `dilationHeight == dilationWidth`. If it is greater than * 1, then all values of `strides` must be 1. * @param strides The strides of the pooling: `[strideHeight, strideWidth]`. If * `strides` is a single number, then `strideHeight == strideWidth`. */ /** @doc {heading: 'Operations', subheading: 'Convolution'} */ function pool_(input, windowShape, poolingType, pad, dilations, strides) { if (dilations == null) { dilations = [1, 1]; } if (strides == null) { strides = 1; } if (pad === 0) { pad = 'valid'; } var $x = tensor_util_env_1.convertToTensor(input, 'x', 'maxPool'); var x4D = $x; var reshapedTo4D = false; if ($x.rank === 3) { reshapedTo4D = true; x4D = $x.as4D(1, $x.shape[0], $x.shape[1], $x.shape[2]); } util.assert(conv_util.eitherStridesOrDilationsAreOne(strides, dilations), 'Error in pool: Either strides or dilations must be 1. ' + ("Got strides " + strides + " and dilations '" + dilations + "'")); var convInfo = conv_util.computePool2DInfo(x4D.shape, windowShape, strides, dilations, pad); var dilation = [convInfo.dilationHeight, convInfo.dilationWidth]; // The following implementation does batchToSpace(pool(spaceToBatch(x))) // whenever dilation > 1 since the TF kernels do not support dilation > 1. // tslint:disable-next-line:max-line-length // https://github.com/tensorflow/tensorflow/blob/50f6bb67dc98c9b74630b6047aae7a4f8a40fd02/tensorflow/python/ops/nn_ops.py#L1037 var basePadding; if (pad === 'same') { basePadding = withSpaceToBatchBasePaddings([convInfo.filterHeight, convInfo.filterWidth], dilation); } else { basePadding = [[0, 0], [0, 0]]; } var isDilationOne = dilation[0] === 1 && dilation[1] === 1; var _a = requiredSpaceToBatchPaddings([convInfo.inHeight, convInfo.inWidth], dilation, basePadding), adjustedPadding = _a[0], adjustedCrops = _a[1]; var convertedPad = isDilationOne ? pad : 'valid'; var convertedX = isDilationOne ? x4D : array_ops_1.spaceToBatchND(x4D, dilation, adjustedPadding); var forwardOp = poolingType === 'avg' ? function () { return avgPoolImpl_(convertedX, windowShape, strides, 1 /* dilation */, convertedPad); } : function () { return maxPoolImpl_(convertedX, windowShape, strides, 1 /* dilation */, convertedPad); }; var y = forwardOp(); var res = isDilationOne ? y : array_ops_1.batchToSpaceND(y, dilation, adjustedCrops); if (reshapedTo4D) { return res.as3D(res.shape[1], res.shape[2], res.shape[3]); } return res; } /** * Computes the backprop of a max pool. * * @param dy The dy error, of rank 4 or rank 3 of shape * [batchSize, height, width, channels]. If rank 3, batch of 1 is * assumed. * @param input The original input image, of rank 4, of shape * [batchSize, height, width, channels]. * @param output The original output image, of rank 4, of shape * [batchSize, outHeight, outWidth, channels]. * @param filterSize The filter size: `[filterHeight, filterWidth]`. If * `filterSize` is a single number, then `filterHeight == filterWidth`. * @param strides The strides of the pooling: `[strideHeight, strideWidth]`. If * `strides` is a single number, then `strideHeight == strideWidth`. * @param pad A string from: 'same', 'valid'. The type of padding algorithm * used in the forward prop of the op. * @param dimRoundingMode A string from: 'ceil', 'round', 'floor'. The * rounding mode used when computing output dimensions if pad is a * number. If none is provided, it will not round and error if the output * is of fractional size. */ function maxPoolBackprop(dy, input, output, filterSize, strides, dilations, pad, dimRoundingMode) { var $dy = tensor_util_env_1.convertToTensor(dy, 'dy', 'maxPoolBackprop'); var $input = tensor_util_env_1.convertToTensor(input, 'input', 'maxPoolBackprop'); var $output = tensor_util_env_1.convertToTensor(output, 'output', 'maxPoolBackprop'); util.assert($input.rank === $dy.rank, "Rank of input (" + $input.rank + ") does not match rank of dy (" + $dy.rank + ")"); if (dilations == null) { dilations = [1, 1]; } util.assert(conv_util.eitherStridesOrDilationsAreOne(strides, dilations), 'Error in maxPoolBackProp: Either strides or dilations must be 1. ' + ("Got strides " + strides + " and dilations '" + dilations + "'")); util.assert($dy.rank === 4, "Error in maxPoolBackprop: dy must be rank 4 but got rank " + ($dy.rank + ".")); util.assert($input.rank === 4, "Error in maxPoolBackprop: input must be rank 4 but got rank " + ($input.rank + ".")); if (dimRoundingMode != null) { util.assert(util.isInt(pad), "Error in maxPoolBackprop: pad must be an integer when using, " + ("dimRoundingMode " + dimRoundingMode + " but got pad " + pad + ".")); } var convInfo = conv_util.computePool2DInfo($input.shape, filterSize, strides, dilations, pad, dimRoundingMode); var res = environment_1.ENV.engine.runKernel(function (backend) { return backend.maxPoolBackprop($dy, $input, $output, convInfo); }, { $dy: $dy, $input: $input }); return res; } /** * Computes the backprop of an avg pool. * * @param dy The dy error, of rank 4 or rank 3 of shape * [batchSize, height, width, channels]. If rank 3, batch of 1 is * assumed. * @param input The input image, of rank 4 or rank 3 of shape * [batchSize, height, width, channels]. If rank 3, batch of 1 is * assumed. * @param filterSize The filter size: `[filterHeight, filterWidth]`. If * `filterSize` is a single number, then `filterHeight == filterWidth`. * @param strides The strides of the pooling: `[strideHeight, strideWidth]`. If * `strides` is a single number, then `strideHeight == strideWidth`. * @param pad A string from: 'same', 'valid'. The type of padding algorithm * used in the forward prop of the op. */ function avgPoolBackprop(dy, input, filterSize, strides, dilations, pad) { var $dy = tensor_util_env_1.convertToTensor(dy, 'dy', 'avgPoolBackprop'); var $input = tensor_util_env_1.convertToTensor(input, 'input', 'avgPoolBackprop'); util.assert($input.rank === $dy.rank, "Rank of input (" + $input.rank + ") does not match rank of dy (" + $dy.rank + ")"); if (dilations == null) { dilations = [1, 1]; } util.assert(conv_util.eitherStridesOrDilationsAreOne(strides, dilations), 'Error in avgPoolBackprop: Either strides or dilations must be 1. ' + ("Got strides " + strides + " and dilations '" + dilations + "'")); var input4D = $input; var dy4D = $dy; var reshapedTo4D = false; if ($input.rank === 3) { reshapedTo4D = true; input4D = $input.as4D(1, $input.shape[0], $input.shape[1], $input.shape[2]); dy4D = $dy.as4D(1, $dy.shape[0], $dy.shape[1], $dy.shape[2]); } util.assert(dy4D.rank === 4, "Error in avgPoolBackprop: dy must be rank 4 but got rank " + (dy4D.rank + ".")); util.assert(input4D.rank === 4, "Error in avgPoolBackprop: input must be rank 4 but got rank " + (input4D.rank + ".")); var convInfo = conv_util.computePool2DInfo(input4D.shape, filterSize, strides, dilations, pad); var res = environment_1.ENV.engine.runKernel(function (backend) { return backend.avgPoolBackprop(dy4D, input4D, convInfo); }, { dy4D: dy4D, input4D: input4D }); if (reshapedTo4D) { return res.as3D(res.shape[1], res.shape[2], res.shape[3]); } return res; } // Helper function to compute crops and paddings for pool with dilation > 1. // tslint:disable-next-line:max-line-length // https://github.com/tensorflow/tensorflow/blob/50f6bb67dc98c9b74630b6047aae7a4f8a40fd02/tensorflow/python/ops/array_ops.py#L2184 function requiredSpaceToBatchPaddings(inputShape, blockShape, basePadding) { var padStart = basePadding.map(function (b) { return b[0]; }); var origPadEnd = basePadding.map(function (b) { return b[1]; }); var fullInputShape = inputShape.concat(padStart, origPadEnd); var padEndExtra = blockShape.map(function (b, i) { return (b - fullInputShape[i] % b) % b; }); var padEnd = origPadEnd.map(function (s, i) { return s + padEndExtra[i]; }); var paddings = blockShape.map(function (_, i) { return [padStart[i], padEnd[i]]; }); var crops = blockShape.map(function (_, i) { return [0, padEndExtra[i]]; }); return [paddings, crops]; } // Helper function to compute base paddings for pool with dilation > 1. // tslint:disable-next-line:max-line-length // https://github.com/tensorflow/tensorflow/blob/50f6bb67dc98c9b74630b6047aae7a4f8a40fd02/tensorflow/python/ops/nn_ops.py#L524 function withSpaceToBatchBasePaddings(filterShape, dilation) { // Spatial dimensions of the filters and the upsampled filters in which we // introduce (rate - 1) zeros between consecutive filter values. var dilatedFilterShape = filterShape.map(function (s, i) { return s + (s - 1) * (dilation[i] - 1); }); var padExtraShape = dilatedFilterShape.map(function (s) { return s - 1; }); // When padding is odd, we pad more at end, following the same // convention as conv2d. var padExtraStart = padExtraShape.map(function (s) { return Math.floor(s / 2); }); var padExtraEnd = padExtraShape.map(function (s, i) { return s - padExtraStart[i]; }); return padExtraShape.map(function (_, i) { return [padExtraStart[i], padExtraEnd[i]]; }); } exports.maxPool = operation_1.op({ maxPool_: maxPool_ }); exports.avgPool = operation_1.op({ avgPool_: avgPool_ }); exports.pool = operation_1.op({ pool_: pool_ }); //# sourceMappingURL=pool.js.map