view src/feature_extractor.hpp @ 0:add35537fdbb tip

Initial import

//  Copyright 2011, Ian Hobson.
//
//  This file is part of gpsynth.
//
//  gpsynth is free software: you can redistribute it and/or modify
//  it under the terms of the GNU General Public License as published by
//  the Free Software Foundation, either version 3 of the License, or
//  (at your option) any later version.
//
//  gpsynth is distributed in the hope that it will be useful,
//  but WITHOUT ANY WARRANTY; without even the implied warranty of
//  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
//  GNU General Public License for more details.
//
//  You should have received a copy of the GNU General Public License
//  along with gpsynth in the file COPYING. 
//  If not, see http://www.gnu.org/licenses/.

// FeatureExtractor - responsible for managing the retrieval of features from 
// audio data. Caches feature data to avoid recalculations when accessing
// different features that have the same dependencies.

#pragma once

#include "mfcc_analyzer.hpp"
#include "spectrum_analyzer.hpp"

#include <complex>
#include <vector>

namespace dsp {

typedef double Value;
typedef std::vector<Value> ValueList;
typedef std::complex<Value> ComplexValue;
typedef std::vector<ComplexValue> ComplexValueList;

class FeatureExtractor {
  std::string file_path_;
  ValueList window_;
  ValueList frame_;
  ValueList file_data_;
  ValueList bin_frequencies_;
  Value bin_size_;
  std::vector<ComplexValueList> spectrum_;
  std::vector<ValueList> magnitude_spectrum_;
  std::vector<ValueList> log_magnitude_spectrum_;
  std::vector<ValueList> mfccs_;
  std::vector<ValueList> delta_mfccs_;
  std::vector<ValueList> double_delta_mfccs_;
  std::vector<ValueList> spectral_flux_;
  ValueList pitch_;
  ValueList spectral_centroid_;
  ValueList spectral_spread_;
  ValueList energy_;
  
  int frames_;
  int window_size_;
  int hop_size_;
  Value sample_rate_;
  
  dsp::SpectrumAnalyzer spectrum_analyzer_;
  dsp::MFCCAnalyzer mfcc_analyzer_;
  
public:
  FeatureExtractor(const std::string& file_path = "",
                   int frame_size = 2048,
                   int hop_size = 512);
  
  void LoadFile(const std::string& file_path);
  void ClearCaches();
  
  int Frames() const { return frames_; }
  
  int WindowSize() const { return window_size_; }
  int HopSize() const { return hop_size_; }
  
  void SetWindowSize(int window_size);
  void SetHopSize(int hop_size);
  
  Value Duration() const;
  
  const std::vector<ComplexValueList>& Spectrum();
  const std::vector<ValueList>& MagnitudeSpectrum();
  const std::vector<ValueList>& LogMagnitudeSpectrum();
  const ValueList& Pitch();
  const ValueList& SpectralCentroid();
  const ValueList& SpectralSpread();
  const std::vector<ValueList>& SpectralFlux();
  const ValueList& Energy();
  const std::vector<ValueList>& MFCCs();
  const std::vector<ValueList>& DeltaMFCCs();
  const std::vector<ValueList>& DoubleDeltaMFCCs();
  
private:
  void GenerateWindow();
};

} // dsp namespace