@tensorflow/tfjs-core/src/backends/cpu/kernels/MaxPoolWithArgmax.ts

Hardware-accelerated JavaScript library for machine intelligence

/**
 * @license
 * Copyright 2020 Google LLC. 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.
 * =============================================================================
 */

import {MaxPoolWithArgmax, MaxPoolWithArgmaxAttrs, MaxPoolWithArgmaxInputs} from '../../../kernel_names';
import {KernelConfig} from '../../../kernel_registry';
import * as conv_util from '../../../ops/conv_util';
import {TypedArray} from '../../../types';
import {MathBackendCPU} from '../backend_cpu';
import {assertNotComplex} from '../cpu_util';

import {maxPoolWithArgmaxImpl} from './MaxPoolWithArgmax_impl';

export const maxPoolWithArgmaxConfig: KernelConfig = {
  kernelName: MaxPoolWithArgmax,
  backendName: 'cpu',
  kernelFunc: ({inputs, attrs, backend}) => {
    const {x} = inputs as MaxPoolWithArgmaxInputs;
    const {filterSize, strides, pad, includeBatchInIndex} =
        attrs as {} as MaxPoolWithArgmaxAttrs;
    const cpuBackend = backend as MathBackendCPU;
    assertNotComplex(x, 'MaxPoolWithArgmax');

    const values = cpuBackend.data.get(x.dataId).values as TypedArray;
    const convInfo = conv_util.computePool2DInfo(
        x.shape as [number, number, number, number], filterSize, strides,
        [1, 1], pad);
    const [pooled, indexes] = maxPoolWithArgmaxImpl(
        values, x.shape, x.dtype, includeBatchInIndex, convInfo);

    const pooledDataId =
        cpuBackend.write(pooled as Float32Array, convInfo.outShape, x.dtype);
    const indexesDataId =
        cpuBackend.write(indexes as Int32Array, convInfo.outShape, x.dtype);
    return [
      {dataId: pooledDataId, shape: convInfo.outShape, dtype: x.dtype},
      {dataId: indexesDataId, shape: convInfo.outShape, dtype: 'int32'}
    ];
  }
};