Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(1272)

Side by Side Diff: webrtc/modules/audio_processing/intelligibility/intelligibility_utils.h

Issue 1693823004: Use VAD to get a better speech power estimation in the IntelligibilityEnhancer (Closed) Base URL: https://chromium.googlesource.com/external/webrtc.git@pow
Patch Set: Created 4 years, 10 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View unified diff | Download patch
OLDNEW
1 /* 1 /*
2 * Copyright (c) 2014 The WebRTC project authors. All Rights Reserved. 2 * Copyright (c) 2014 The WebRTC project authors. All Rights Reserved.
3 * 3 *
4 * Use of this source code is governed by a BSD-style license 4 * Use of this source code is governed by a BSD-style license
5 * that can be found in the LICENSE file in the root of the source 5 * that can be found in the LICENSE file in the root of the source
6 * tree. An additional intellectual property rights grant can be found 6 * tree. An additional intellectual property rights grant can be found
7 * in the file PATENTS. All contributing project authors may 7 * in the file PATENTS. All contributing project authors may
8 * be found in the AUTHORS file in the root of the source tree. 8 * be found in the AUTHORS file in the root of the source tree.
9 */ 9 */
10 10
11 #ifndef WEBRTC_MODULES_AUDIO_PROCESSING_INTELLIGIBILITY_INTELLIGIBILITY_UTILS_H_ 11 #ifndef WEBRTC_MODULES_AUDIO_PROCESSING_INTELLIGIBILITY_INTELLIGIBILITY_UTILS_H_
12 #define WEBRTC_MODULES_AUDIO_PROCESSING_INTELLIGIBILITY_INTELLIGIBILITY_UTILS_H_ 12 #define WEBRTC_MODULES_AUDIO_PROCESSING_INTELLIGIBILITY_INTELLIGIBILITY_UTILS_H_
13 13
14 #include <complex> 14 #include <complex>
15 #include <vector>
15 16
16 #include "webrtc/base/scoped_ptr.h" 17 #include "webrtc/base/scoped_ptr.h"
17 18
18 namespace webrtc { 19 namespace webrtc {
19 20
20 // Internal helper for computing the power of a stream of arrays. 21 // Internal helper for computing the power of a stream of arrays.
21 // The result is an array of power per position: the i-th power is the power of 22 // The result is an array of power per position: the i-th power is the power of
22 // the stream of data on the i-th positions in the input arrays. 23 // the stream of data on the i-th positions in the input arrays.
23 class PowerEstimator { 24 class PowerEstimator {
24 public: 25 public:
25 // Construct an instance for the given input array length (|freqs|), with the 26 // Construct an instance for the given input array length (|freqs|), with the
26 // appropriate parameters. |decay| is the forgetting factor. 27 // appropriate parameters. |decay| is the forgetting factor.
27 PowerEstimator(size_t freqs, float decay); 28 PowerEstimator(size_t freqs, float decay);
28 29
29 // Add a new data point to the series. 30 // Add a new data point to the series.
31 void Step(const float* data);
30 void Step(const std::complex<float>* data); 32 void Step(const std::complex<float>* data);
31 33
32 // The current power array. 34 // The current power array.
33 const float* Power(); 35 const float* power() { return &power_[0]; };
hlundin-webrtc 2016/02/15 13:05:12 Why not return a const std::vector<float>&?
aluebs-webrtc 2016/02/19 03:56:31 Done.
34 36
35 private: 37 private:
36 // TODO(ekmeyerson): Switch the following running means 38 // The current power array.
37 // and histories from rtc::scoped_ptr to std::vector. 39 std::vector<float> power_;
38 rtc::scoped_ptr<std::complex<float>[]> running_mean_sq_;
39 40
40 // The current magnitude array.
41 rtc::scoped_ptr<float[]> magnitude_;
42 // The current power array.
43 rtc::scoped_ptr<float[]> power_;
44
45 const size_t num_freqs_;
46 const float decay_; 41 const float decay_;
47 }; 42 };
48 43
49 // Helper class for smoothing gain changes. On each application step, the 44 // Helper class for smoothing gain changes. On each application step, the
50 // currently used gains are changed towards a set of settable target gains, 45 // currently used gains are changed towards a set of settable target gains,
51 // constrained by a limit on the magnitude of the changes. 46 // constrained by a limit on the magnitude of the changes.
52 class GainApplier { 47 class GainApplier {
53 public: 48 public:
54 GainApplier(size_t freqs, float change_limit); 49 GainApplier(size_t freqs, float change_limit);
55 50
56 // Copy |in_block| to |out_block|, multiplied by the current set of gains, 51 // Copy |in_block| to |out_block|, multiplied by the current set of gains,
57 // and step the current set of gains towards the target set. 52 // and step the current set of gains towards the target set.
58 void Apply(const std::complex<float>* in_block, 53 void Apply(const std::complex<float>* in_block,
59 std::complex<float>* out_block); 54 std::complex<float>* out_block);
60 55
61 // Return the current target gain set. Modify this array to set the targets. 56 // Return the current target gain set. Modify this array to set the targets.
62 float* target() const { return target_.get(); } 57 float* target() const { return target_.get(); }
63 58
64 private: 59 private:
65 const size_t num_freqs_; 60 const size_t num_freqs_;
66 const float change_limit_; 61 const float change_limit_;
67 rtc::scoped_ptr<float[]> target_; 62 rtc::scoped_ptr<float[]> target_;
68 rtc::scoped_ptr<float[]> current_; 63 rtc::scoped_ptr<float[]> current_;
69 }; 64 };
70 65
71 } // namespace webrtc 66 } // namespace webrtc
72 67
73 #endif // WEBRTC_MODULES_AUDIO_PROCESSING_INTELLIGIBILITY_INTELLIGIBILITY_UTILS _H_ 68 #endif // WEBRTC_MODULES_AUDIO_PROCESSING_INTELLIGIBILITY_INTELLIGIBILITY_UTILS _H_
OLDNEW

Powered by Google App Engine
This is Rietveld 408576698