face-api/src/faceFeatureExtractor/TinyFaceFeatureExtractor.ts

import * as tf from '../../dist/tfjs.esm';

import { NetInput, TNetInput, toNetInput } from '../dom/index';
import { NeuralNetwork } from '../NeuralNetwork';
import { normalize } from '../ops/index';
import { denseBlock3 } from './denseBlock';
import { extractParamsFromWeightMapTiny } from './extractParamsFromWeightMapTiny';
import { extractParamsTiny } from './extractParamsTiny';
import { IFaceFeatureExtractor, TinyFaceFeatureExtractorParams } from './types';

export class TinyFaceFeatureExtractor extends NeuralNetwork<TinyFaceFeatureExtractorParams> implements IFaceFeatureExtractor<TinyFaceFeatureExtractorParams> {
  constructor() {
    super('TinyFaceFeatureExtractor');
  }

  public forwardInput(input: NetInput): tf.Tensor4D {
    const { params } = this;

    if (!params) {
      throw new Error('TinyFaceFeatureExtractor - load model before inference');
    }

    return tf.tidy(() => {
      const batchTensor = tf.cast(input.toBatchTensor(112, true), 'float32');
      const meanRgb = [122.782, 117.001, 104.298];
      const normalized = normalize(batchTensor, meanRgb).div(255) as tf.Tensor4D;

      let out = denseBlock3(normalized, params.dense0, true);
      out = denseBlock3(out, params.dense1);
      out = denseBlock3(out, params.dense2);
      out = tf.avgPool(out, [14, 14], [2, 2], 'valid');

      return out;
    });
  }

  public async forward(input: TNetInput): Promise<tf.Tensor4D> {
    return this.forwardInput(await toNetInput(input));
  }

  protected getDefaultModelName(): string {
    return 'face_feature_extractor_tiny_model';
  }

  protected extractParamsFromWeightMap(weightMap: tf.NamedTensorMap) {
    return extractParamsFromWeightMapTiny(weightMap);
  }

  protected extractParams(weights: Float32Array) {
    return extractParamsTiny(weights);
  }
}