ente/thirdparty/face-api/faceRecognitionNet/residualLayer.ts

40 lines
1.3 KiB
TypeScript
Raw Normal View History

import * as tf from '@tensorflow/tfjs-core';
import { conv, convDown, convNoRelu } from './convLayer';
import { ResidualLayerParams } from './types';
export function residual(x: tf.Tensor4D, params: ResidualLayerParams): tf.Tensor4D {
let out = conv(x, params.conv1)
out = convNoRelu(out, params.conv2)
out = tf.add(out, x)
out = tf.relu(out)
return out
}
export function residualDown(x: tf.Tensor4D, params: ResidualLayerParams): tf.Tensor4D {
let out = convDown(x, params.conv1)
out = convNoRelu(out, params.conv2)
let pooled = tf.avgPool(x, 2, 2, 'valid') as tf.Tensor4D
const zeros = tf.zeros<tf.Rank.R4>(pooled.shape)
const isPad = pooled.shape[3] !== out.shape[3]
const isAdjustShape = pooled.shape[1] !== out.shape[1] || pooled.shape[2] !== out.shape[2]
if (isAdjustShape) {
const padShapeX = [...out.shape] as [number, number, number, number]
padShapeX[1] = 1
const zerosW = tf.zeros<tf.Rank.R4>(padShapeX)
out = tf.concat([out, zerosW], 1)
const padShapeY = [...out.shape] as [number, number, number, number]
padShapeY[2] = 1
const zerosH = tf.zeros<tf.Rank.R4>(padShapeY)
out = tf.concat([out, zerosH], 2)
}
pooled = isPad ? tf.concat([pooled, zeros], 3) : pooled
out = tf.add(pooled, out) as tf.Tensor4D
out = tf.relu(out)
return out
}