Voice privacy system and/or associated method
阅读说明：本技术 语音隐私系统和/或相关联的方法 (Voice privacy system and/or associated method ) 是由 阿列克谢·克拉斯诺夫 于 2018-03-14 设计创作，主要内容包括：本发明的某些示例性实施方案涉及语音隐私系统和/或相关联的方法。本文所述的技术通过例如将原始语音信号的屏蔽副本叠加到该原始语音信号上来干扰感知的语音的可理解性,其中该信号的某些部分通过时间延迟和/或振幅调节来模糊,其中该时间延迟和/或振幅调节随时间推移而振荡。在某些示例性实施方案中,可在与话音素、辅音声音、音素、以及/或者语音的其他相关或不相关的信息承载的语音构建块对应的频率范围内生成原始信号的模糊。除此之外或作为另外一种选择,可将特定于房间或区域的低频范围内的烦扰性混响从副本信号中“切除”,而不增大或不显著增大感知的响度。(Certain example embodiments of this invention relate to voice privacy systems and/or associated methods. The techniques described herein interfere with the intelligibility of perceived speech by, for example, superimposing a masked copy of the original speech signal onto the original speech signal, where certain portions of the signal are obscured by time delays and/or amplitude adjustments that oscillate over time. In some exemplary embodiments, ambiguities of the original signal may be generated in frequency ranges corresponding to the phonemes of phonemes, consonantal sounds, phonemes, and/or other related or unrelated information-bearing speech building blocks of speech. Additionally or alternatively, disturbing reverberation in the low frequency range specific to a room or region can be "cut out" from the replica signal without increasing or not significantly increasing the perceived loudness.)
1. A method for disturbing speech intelligibility, the method comprising:
receiving an original speech signal corresponding to an original speech via a microphone;
generating an intelligible interference masking signal comprising the blurred cues of the original speech in the original speech signal; and
reducing the level of intelligibility of the original speech signal by outputting the intelligibility interference masking signal comprising the blurred speech prompt via a loudspeaker.
2. The method according to claim 1, wherein said intelligibility interference mask signal is time delayed with respect to said original speech signal.
3. The method according to claim 2 wherein the intelligible interference mask signal is time delayed by 20ms to 150 ms.
4. The method according to any of claims 2 to 3 wherein the intelligible interference masking signal is time delayed by 80 ms.
5. The method of any of claims 2 to 4, wherein the time delay oscillates over time.
6. The method of claim 5, wherein the time delay oscillates in time in a range of 80ms to 230ms with respect to the original signal.
7. The method of any one of claims 2 to 6, wherein the time delay oscillates at an oscillation frequency of 2Hz to 6 Hz.
8. The method of any one of claims 2 to 6, wherein the time delay oscillates at an oscillation frequency of 1Hz to 10 Hz.
9. The method according to any of the preceding claims, wherein the intelligible interference mask signal is generated such that the amplitude of the intelligible interference mask signal is modulated over time.
10. The method according to any of the preceding claims, further comprising generating the intelligibility interference mask signal such that the gain corresponding to the intelligibility interference mask signal added to the original speech signal is between 0.05% and 0.25%.
11. A method according to any preceding claim, wherein the ambiguous prompts are generated at a frequency of 0.01Hz to 20 Hz.
12. The method of claim 11, wherein the ambiguous prompts are generated at a frequency of 2Hz to 6 Hz.
13. The method according to any of the preceding claims, further comprising outputting the intelligibility interference masking signal and a pre-recorded mix of multiple voices through the loudspeaker.
14. The method of claim 13, wherein the pre-recorded mix of multiple voices includes 2 to 7 different voices.
15. The method of claim 13, wherein the pre-recorded mix of multiple voices includes 3 different voices.
16. A speech intelligibility interfering device comprising:
a control circuit configured to:
receiving an original voice signal corresponding to an original voice from a microphone;
generating an intelligible interference masking signal comprising the blurred cues of the original speech in the original speech signal; and
causing a speaker to output the intelligibility-interference-masked signal including the blurred voice prompt to reduce an intelligibility level of the original voice signal.
17. The apparatus according to claim 16, wherein said intelligibility-interference-masked signal is time-delayed with respect to said original speech signal.
18. The apparatus according to claim 17 wherein said intelligible interference mask signal is time delayed by 20ms to 150 ms.
19. The apparatus of any of claims 17-18, wherein the time delay oscillates over time.
20. The apparatus of any of claims 17-19, wherein the time delay oscillates in time in a range of 80ms to 230ms with respect to the original signal.
21. The apparatus of any one of claims 17 to 20, wherein the time delay oscillates at an oscillation frequency of 1Hz to 10 Hz.
22. The apparatus according to any of the claims 16 to 21 wherein the intelligible interference mask signal is generated such that the amplitude of the intelligible interference mask signal is modulated over time.
23. The apparatus according to any of the claims 16 to 22, wherein the intelligibility interference mask signal is generated such that the gain corresponding to the intelligibility interference mask signal added to the original speech signal is between 0.05% and 0.25%.
24. The device of any of claims 16 to 23, wherein the ambiguous prompts are generated at a frequency of 2Hz to 6 Hz.
25. The apparatus according to any of the claims 16 to 24 further comprising outputting the intelligible interference masking signal through the speaker and a pre-recorded mix of a plurality of voices.
26. A speech intelligibility interference system comprising:
a microphone;
a speaker; and
a control circuit configured to:
receiving an original speech signal corresponding to original speech from the microphone;
generating an intelligible interference masking signal comprising the blurred cues of the original speech in the original speech signal; and
causing the speaker to output the intelligibility-interference-masked signal including the blurred voice prompt to reduce an intelligibility level of the original voice signal.
27. The system according to claim 26, wherein said intelligibility-interference-masked signal is time-delayed relative to said original speech signal, said time-delay oscillating over time.
28. The system according to any one of claims 26 to 27 wherein the understandability interference mask signal is generated such that the amplitude of the understandability interference mask signal oscillates over time.
29. An acoustic wall comprising the system of any of claims 26-28.
Background and summary of the invention
Protecting voice privacy has become an increasingly important task in modern workplaces. Speakers want their speech content limited to their office or conference room. On the other hand, unintended listeners do not want to be disturbed by unnecessary verbal information. In environments other than offices, including, for example, homes, libraries, banks, etc., where people are often unaware that their speech interferes with others, irritating speech from others can also be problematic.
In fact, a persistent disturbing sound can cause a variety of potential negative effects. These negative effects can range from loss of productivity of the tissue (e.g., inability to sustain and/or interrupt, concentrate) to human medical problems (e.g., headache episodes due to disturbing sounds, stress, increased heart rate, etc.), to even an intense desire to find new work environments. Phonophobia is a habitual disorder related to the association of sounds with unpleasant things, which also occurs from time to time. Some people exhibit excessive alertness or hypersensitivity to certain sounds and invasive speech.
In many environments, sound annoyance is often related to loudness, abruptness, high pitch, and in the case of speech sounds, also to speech content. In many cases, certain components in speech or noise make it particularly disturbing or irritating. For speech content, people tend to try to hear the content, regardless of the volume, which has been found to increase the annoyance subconsciously. That is, once a person is aware that someone is speaking, it is often involuntarily involved, thereby adding a subconscious annoyance.
People are often stimulated with high frequencies (e.g., sounds in the range of 2,000-4,000 Hz). These sounds need not have high intensity to be perceived as loud. In this regard, fig. 1 is a graph showing human hearing perceived at a constant level, plotting sound pressure level versus frequency. As shown, the "equal loudness sound curve" in fig. 1 shows that low frequency sounds with high sound pressure levels are generally perceived in the same way as higher frequency sounds with lower sound pressure levels. In general, the stimulus increases with the volume of the noise.
Sound waves (including speech) propagate in a longitudinal fashion primarily through alternating compressions and rarefactions of air. When a wave strikes a wall, the deformation of the molecules creates pressure on the outside of the wall, which in turn creates a secondary sound.
It will be appreciated that it would be desirable to design the wall with noise cancellation (including voice interference characteristics) for at least some environments. Some building materials (including glass) are poor sound insulators. At the same time, the use of glass is often advantageous because it provides excellent visual connectivity between offices and can aid in employee involvement. It will therefore be appreciated that it would be desirable to design optically transparent walls with noise cancellation properties (including voice interference properties) for at least some of these environments.
Acoustical windows are known in the art. One mainstream approach involves increasing the Sound Transmission Class (STC) of the wall. STC is an integer rating of the degree of sound attenuation of a wall. It is weighted against 16 frequencies over the entire human hearing range. The STC can be increased, for example, by: using a specific spacing in combination with a double glass wall in order to resourcefully resonate sound; the STC of single or double walls is increased by increasing the thickness of the glass and/or using laminated glass.
Unfortunately, however, these techniques come at a cost. For example, increasing the thickness of a single ply of glass allows only moderate sound attenuation, while also increasing cost. The use of double glazing, while more effective, generally requires the use of at least two thicker (e.g., 6-12.5mm) sheets of glass. These methods also generally require high tolerances in the wall construction and the use of special pliable mechanical connectors to avoid the flanking effect. Glass of such thickness is heavy and expensive and can result in high installation costs.
In addition, double walls are generally suitable for low frequency sounds in general. This limits their effectiveness to a smaller number of applications, such as being confined to the outer wall to counteract the low frequency noise of jet and automobile engines, harbour, rail, etc. At the same time, most speech sounds that cause annoyance and speech recognition are in the 1800+ Hz range. It would therefore be desirable to implement noise cancellation in this higher frequency range, for example, to help block the irritating components and improve voice privacy.
Some acoustic solutions focus on sound masking rather than reducing higher frequency noise. For example, sounds of various frequencies may be electronically overlapped by speakers such that additional sound is provided "above the original noise. Sound masking may include natural sounds, ranging from waterfalls and rain sounds to fire pop sounds and thunderstorm sounds. Various types of artificially generated masking noise, such as white noise, pink noise, brown noise, and other noise, are also used in this regard. The main purpose of these sound masking techniques relates to reducing the annoyance of the surrounding noise, and such methods are indeed able to mask the irritancy. Unfortunately, however, this also creates additional noise, which some people may perceive as irritating itself. One problem with the above-described sound masking techniques is that their frequencies lie outside the range of frequencies of occurrence of syllables (building blocks of speech). See, for example, fig. 11, discussed in more detail below, which shows the results of a time-frequency analysis of a normal speech pattern, white noise, and some natural sound maskers.
Yet another exemplary method for implementing noise cancellation is used in Bose headphones. The method involves registering incoming noise and generating cancellation noise out of phase with the registered incoming noise. While it is easier to isolate a person from the environment by wearing the headset, doing so does not prevent the person wearing the headset from making noise that is perceived by others as disturbing. That is, even though the person wearing the headphones may create an isolated environment on a personal level, there is still a problem in creating an isolated area for the group so that others in the group cannot hear the content. Furthermore, one difficulty with the wall of this concept is that it is generally applicable only to small areas and primarily to continuous low frequency sounds (such as the booming sound of an engine). One reason for this is that only narrow frequency bands can be effectively tuned out of phase and the higher the frequency, the smaller the auditory space for effective noise cancellation.
It will thus be appreciated that it would be desirable to provide techniques that overcome some or all of the above and/or other voice masking problems. For example, it will be appreciated that it would be desirable to provide acoustic techniques that help reduce or otherwise compensate for sounds (including speech) that cause irritation and annoyance to people.
The inventors of the present invention have recognized that it would be desirable to prevent the content of speech from being understood by people around speakers in environments such as open or closed office spaces and/or other environments, adjacent offices separated by thin walls with low STC, vehicles (including, for example, commercial and private cars, such as cars, trucks, trains, planes, etc.), bank teller office spaces, hospitals, police stations, conference rooms, etc. In fact, in a broad sense, in modern office spaces, the requirements in terms of sound privacy seem to be increasing.
Current techniques, including the sound masking and sound cancellation techniques discussed above, are not targeted at the content of speech, and in particular are not speech intelligibility interference techniques. Indeed, noise masking techniques known in the art are not fundamentally intended to effectively interfere with speech without causing a significant amount of additional annoyance. In this regard, the present inventors have recognized that while the fundamental frequency of human speech does lie in the same frequency spectrum as some of the available masking noise and/or range that may be at least partially eliminated, it has been found that blocks containing information occur at substantially different frequencies. The blocks containing information in this context are the phonemes representing bursts of sound energy.
It has therefore been recognized that it would be desirable to develop sound masking techniques that are intended to interfere with the informational content of speech without causing additional annoyance. It should be appreciated that masking techniques typically add a certain amount of loudness above the original speech. The techniques of certain example embodiments add only a small amount of extra loudness, for example, because they specifically target basic cues of speech (such as phonemes).
In certain exemplary embodiments, a method for disturbing speech intelligibility is provided, the method comprising: receiving an original speech signal corresponding to an original speech via a microphone; generating an intelligibility interference masking signal containing the fuzzy cues of the original speech in the original speech signal; and reducing the level of intelligibility of the original speech signal by outputting an intelligibility interference masking signal comprising the blurred speech prompt via the loudspeaker.
Devices and systems incorporating such functionality, and walls incorporating such devices and systems are also contemplated herein.
Features, aspects, advantages, and example embodiments described herein may be combined to realize another embodiment.
Drawings
These and other features and advantages may be better and more completely understood by reference to the following detailed description of exemplary illustrative embodiments in conjunction with the accompanying drawings, of which:
FIG. 1 is a graph showing human hearing perceived at a constant level, plotting sound pressure level versus frequency;
fig. 2 is a schematic diagram of some examples of situations occurring at different reverberation times and showing exemplary applications applicable to the different reverberation times;
FIG. 3 shows the T calculated in a variable size room with walls made of three different materials (i.e., glass, polycarbonate, and grey board)60；
Fig. 4A-4B provide examples of the effects that reverberation may have;
FIG. 5 is a graph depicting STC and T according to certain example embodiments60The graph of (a) further identifies some of the advantages that arise when using active methods for speech intelligibility interference;
6A-6B are schematic diagrams of an acoustic wall assembly incorporating an active noise speech intelligibility interference method according to certain example embodiments;
FIG. 7 is a schematic diagram of another acoustic wall assembly incorporating an active speech intelligibility interference method according to some example embodiments;
8A-8B are schematic diagrams of an acoustic wall assembly incorporating an active speech intelligibility interference method that may be used in conjunction with two walls, according to some example embodiments;
FIG. 9 is a flow chart illustrating an exemplary method for active speech intelligibility interference, which may be used in connection with certain exemplary embodiments;
FIG. 10 shows the voiced pixel frequencies for single-voiced speech and multi-voiced speech at their top and bottom portions, respectively;
FIG. 11 illustrates the phoneme frequencies of different types of sounds, including different natural sounds and different sounds of speech; and is
FIG. 12 is a block diagram of an electronic voice intelligibility interfering device according to certain example embodiments;
FIG. 13 includes examples of the frequency dependence of various syllables, wherein each syllable includes a consonant and a vowel;
fig. 14 is a block diagram of an electronic device that facilitates reducing disturbing reverberation in a room, according to some example embodiments;
FIG. 15 is a graph showing an exemplary masking signal (gray) superimposed on an original speech signal (black); and is
Fig. 16 provides test data obtained from samples made according to certain exemplary embodiments.
Detailed Description
Certain example embodiments relate to acoustic wall assemblies that use active (via electronic means) acoustic reverberation to achieve speech intelligibility interference functions, and/or methods of making and/or using the same. The actively added reverberation helps to mask the irritating sounds originating from inside or outside the room provided with such wall components. In certain exemplary embodiments, such methods include, for example, helping to make otherwise potentially interfering speech perceived as unintelligible (and, therefore, less annoying).
Certain exemplary embodiments add noise shielding and voice interference characteristics to walls with low STC, advantageously allowing for low cost, low weight solutions with voice privacy quality. Certain exemplary embodiments can be used in high STC walls, for example, as a means to further improve voice privacy and/or noise shielding.
Reverberation is sometimes advantageous compared to conventional sound deadening and masking techniques. For example, reverberation in some cases only increases the loudness necessary to interfere with speech or noise. In some implementations, no or only minimal unnecessary additional noise is generated. Reverberation is also advantageously not limited to a particular wall assembly size and/or geometry, works equally well at low and high frequencies, and is "forgiving" to the presence of side wing losses (which otherwise sometimes impede sound isolation by sound vibrations passing through structures along the incident path, such as through frame connections, electrical outlets, embedded lights, water pipes, ductwork, and other acoustic gaps). Reverberation is also advantageously resistant to surveillance. Speech masked by white noise can sometimes be prone to deciphering (e.g., by removing additional randomly generated noise from the signal), while reverberation is difficult to decode because there is substantially no reference signal (e.g., it is substantially self-referencing). Furthermore, in at least some cases, reverberation is activated by the original speech signal and its volume is automatically adjusted to follow the volume of the original signal. A further benefit of using reverberation relates to its ability to interfere with so-called "hits", which are potentially irritating low frequency sounds made up of two different sound frequencies. Although low frequency sounds may not always be heard, they may themselves have undesirable subconscious effects. Furthermore, reverberation may be advantageous from a cost perspective because it only interferes with the informative portion of speech, rather than attempting to completely cover speech at the expense of loudness. In practice, the energy required for reverberation will often be less than that required to add white noise.
In particular, when speech is involved, certain exemplary embodiments are effective in: the rhythm of interfering speech, including the fundamental frequencies and their harmonics; masking key voice cues for overlapping syllables and vowels; eliminating artificially generated low frequency sounds having a sub-threshold frequency that produces an adverse resonance with brain waves; and so on. Certain exemplary embodiments use reverberation in the 4-6Hz range, which corresponds to the number of syllables pronounced per second in normal english speech.
Reverberation time T60Is a measure associated with reverberation. It represents the time required for a sound to decay 60 decibels from its initial level. Rooms with different purposes benefit from different reverberation times. Fig. 2 is a schematic diagram of some examples of situations occurring at different reverberation times and illustrates an exemplary application applicable to different reverberation times. Generally, too low a T60Values (e.g., low to no reverberation) tend to "dry out" the speech sounds and are preferred in conference rooms, classrooms, and offices, while too high a T60Values (e.g., providing a lot of reverberation) tend to make speech richer and used in concert halls, churches, and the like. The very high value of T60 makes speech unintelligible.
T60May be calculated based on the race formula:
in this formula, V is the volume, SeIs the combined effective surface area of the room. Calculating S for each wall by multiplying the physical area by the sound absorption coefficienteThe absorption coefficient is a teaching value that varies depending on the material. The following table provides the sound absorption of some commonly used interior building materialsAnd (4) the coefficient.
Flooring material
125Hz
250Hz
500Hz
1kHz
2kHz
4kHz
Foam carpet
0.08
0.24
0.57
0.69
0.71
0.73
Wall material
125Hz
250Hz
500Hz
1kHz
2kHz
4kHz
Brick making: unglazed
0.03
0.03
0.03
0.04
0.05
0.07
Curtain: 10 ounces per square yard
0.03
0.04
0.11
0.17
0.24
0.35
Glass fiber: 2"
0.17
0.55
0.80
0.90
0.85
0.80
Glass: 1/4' board is big
0.18
0.06
0.04
0.03
0.02
0.02
Polycarbonate resin
0.27
0.38
0.25
0.18
0.1
0.07
Ceiling material
125Hz
250Hz
500Hz
1kHz
2kHz
4kHz
Sound-absorbing ceiling inlaid block
0.70
0.66
0.72
0.92
0.88
0.75
3/8' plywood panel
0.28
0.22
0.17
0.09
0.10
0.11
FIG. 3 shows the T calculated in a variable size room with walls made of three different materials (i.e., glass, polycarbonate, and grey board)60。
Examples of the effects that reverberation may have are shown in fig. 4A-4B. Fig. 4A represents the original speech pattern, and fig. 4B illustrates exemplary effects that reverberation may have. As can be seen from fig. 4A-4B, reverberation interferes with speech intelligibility by (among other things) filling the "spaces" between the phonemes of a voice, which can be considered a cluster of acoustic energy. Adding signals to these speech building blocks (i.e., vowels and (especially) consonants) and disturbing the spaces between phonemes helps to make the speech unintelligible and reduce the potentially adverse psychoacoustic effects of the speech.
As described above, certain exemplary embodiments may use active methods to trigger reverberation for noise masking and speech intelligibility interference effects. As will become more apparent from the following description, active methods may involve electronic, electromechanical, and/or selectively controllable mechanical devices to interfere with acoustic waves incident on and/or near wall assemblies and the like. In certain exemplary embodiments, passive methods may supplement such techniques. In this regard, passive methods may involve, for example, wall assemblies specifically designed to trigger reverberation, e.g., attaching or otherwise forming acoustic reverberation components in and/or on the wall assemblies by incorporating holes in the wall assemblies and/or using the natural characteristics of the walls themselves formed thereby, etc.
Referring again to fig. 3, it can be seen that the reverberation in the wall is predominantly noticeable in the low frequency range. Thus, in some cases, it may be desirable to use an active approach in order to mask the informative content of the irritating sounds and speech using reverberation in the high frequency range. Fig. 5 is a graph depicting STC versus T60, further identifying some advantages that result when using an active approach for speech intelligibility interference, according to certain exemplary embodiments. That is, as can be seen in fig. 5, when processing low T60 values, a high STC may be required to make voice and/or similar content unintelligible. In contrast, electronically created mechanisms may help make the perceived speech unintelligible even at low STC values.
Fig. 6A is a schematic diagram of an acoustic wall assembly incorporating an active speech intelligibility interference method according to some example embodiments. As shown in fig. 6A,
The
In response to detecting sound waves within one or more predetermined frequency ranges,
Although in fig. 6A, the
In addition to or instead of reverberation, certain exemplary embodiments may implement active masking by inverse masking. The noise masking enabled by the
The
For a cross-sectional view, the outer
As mentioned above, the wall may be made of or comprise glass. That is, certain example embodiments may relate to a glass wall used in conjunction with an acoustic wall assembly. The glass wall may comprise one, two, three or another number of glass sheets. The glass may be regular float glass, heat strengthened glass, tempered glass, and/or laminated glass. In certain exemplary embodiments, the wall may be composed of or include an Insulating Glass (IG) unit, a Vacuum Insulating Glass (VIG) unit, and the like. The IG unit can include first and second substantially parallel spaced apart substrates, with an edge seal formed around the peripheral edge, and with the cavity between the substrates optionally filled with an inert gas (e.g., Ar, Xe, etc.) with or without air. The VIG unit may include: a first substrate and a second substrate spaced apart substantially in parallel, wherein an edge seal is formed around a peripheral edge; and a gasket, wherein the cavity between the substrates is vented to a pressure below atmospheric. In some cases, a frame may be provided around the IG unit and/or the VIG unit, and may be part of the acoustic wall assembly. In certain exemplary embodiments, other transparent materials may be used. In certain exemplary embodiments, the natural high acoustic reflection coefficient of the glass may be advantageous, for example, when reverberation and/or other noise masking effects are triggered.
Fig. 6B is similar to fig. 6A, except that first and second microphones 606A and 606B are provided so that
FIG. 7 is a schematic diagram of another acoustic wall assembly incorporating an active speech intelligibility interference method according to some example embodiments. Fig. 7 shows a
In some embodiments, one or more speakers may be positioned outside of
Fig. 8A-8B are schematic diagrams of an acoustic wall assembly incorporating an active speech intelligibility interference method that may be used in conjunction with two walls, according to some example embodiments. Fig. 8A to 8B are similar to fig. 6A to 6B. However, rather than having a single wall outer and inner surfaces, outer and
It is believed that the lateral dimensions of the wall may primarily affect the fundamental spectral region of speech and its lower harmonics, while the distance between the two plates of the wall will primarily affect the high frequency content and its higher harmonics. Exemplary embodiments of the glass wall have dimensions of 10ft. × 12ft., with the air spacing between the two glass sheets preferably being in the range of 1-20cm, more preferably in the range of 7-17cm, and an exemplary spacing of 10 cm.
FIG. 9 is a flow chart illustrating an exemplary method for active speech intelligibility interference that may be used in connection with certain exemplary embodiments. Fig. 9 assumes that a wall or wall assembly has been provided (step S902). An incident sound wave is detected (step S904). If the detected sound wave is not within or does not include the frequency range of interest (as determined in step S906), the process simply returns to step S904 and waits for additional incident sound waves to be detected. On the other hand, if the detected sound waves are within or include the frequency range of interest (as determined in step S906), the speaker is used to generate a speech intelligibility interference signal (step S908), for example, according to an exemplary algorithm discussed in more detail below. This behavior thus provides dynamic or "on-demand" masking of noise, including, for example, interfering with speech intelligibility through systems that are not always "on". If the voice is not terminated (as determined in step S910), the process returns to step S908 and a speech intelligibility interference signal is still generated. On the other hand, if the sound is terminated, information about the event may be recorded (step S912), and the process may return to step S904 and wait for the detection of additional incident sound waves.
The recording of step S912 may include, for example, creating a record in a data file stored to a non-transitory computer readable storage medium or the like (e.g., flash memory, USB drive, RAM, etc.). The record may include a timestamp indicating the start time and stop time of the event, as well as a location identifier (e.g., a wall that detects sound if, for example, there are multiple walls implementing the techniques disclosed herein, a microphone that detects sound if, for example, there are multiple microphones in a given wall, etc.). Information relating to the detected and/or generated frequency ranges and/or signals may also be stored to a record. In certain example embodiments, the circuitry may store a digital or other representation of the detected and/or generated sound, for example, in a record or associated data file. Thus, voice or other noise may be recorded, where the entire conversation may be captured and archived for potential subsequent analysis. For example, the sound masking circuit may be used, for example, as a recording device (e.g., a security camera, an eavesdropping device, a sound statistics monitoring device, etc.). In certain example embodiments, information may be stored locally and/or transmitted to a remote computer terminal or the like for potential follow-up actions, such as playing back noise events and/or conversations, analyzing noise events and/or conversations (e.g., to help reveal what types of noise are primarily recorded, when during the day it is loud, who makes the most of the different noises, etc.). Transmission may be accomplished through removable physical media (such as a flash drive, USB drive, etc.), via a wired connection (e.g., including transmission through serial, USB, ethernet, or other cable), wirelessly (e.g., through Wi-Fi, bluetooth, through the internet, etc.), and so forth. In various exemplary embodiments, information may be transmitted periodically and/or on demand.
In certain exemplary embodiments, the sound masking circuit may be programmed to determine whether the incident noise corresponds to a known pattern or type. For example, alarm sounds, sirens, etc., while disturbing, may be detected by the sound masking circuit and allowed to pass through the wall assembly for security, informational, and/or other purposes.
In certain example embodiments, the sound masking circuit may be programmed to operate as both a sound (e.g., speech) jammer (e.g., by using reverberation, etc.) and a sound beautifier. For the latter, the sound masking circuit may generate reverberation and/or pleasant sound to help mask potentially disturbing noise and/or interfere with the intelligibility of speech. The pleasant sound may be a natural sound (e.g., the sound of the sea, a lightning strike, rain, a waterfall, etc.), the sound of an animal (e.g., a dolphin), soothing music, etc. These sounds may be stored in a data memory accessible to the sound masking circuit. The sound masking circuit may retrieve the sound beautifier and provide it as output to a speaker or the like (which may be, for example, the same or a different speaker as used as the air pump in certain exemplary embodiments) as appropriate (e.g., when reverberation is triggered as described above).
It should be appreciated that passive methods for noise interference and/or cancellation may be used in certain exemplary embodiments, for example, because the walls themselves may be configured to act as reverberation-inducing resonators related to acoustic contrast. This can be achieved by: one or more (and preferably two or more) openings, slits, etc. are formed in the acoustic wall assembly to use the natural characteristics of the wall itself to produce the desired type of reverberation effect. These features may be formed on one side of the acoustic wall assembly, thereby adding directional characteristics to the acoustic effect of the wall assembly. For example, at least one opening may be formed in the outer layer of the double wall in order to make the effect directional and make the reverberation effect more pronounced outside the wall. As another example, at least one opening may be formed in the double-walled inner layer. This may be advantageous for some applications, such as concert halls, which may benefit from additional sound reverberation making the sound appear richer.
In certain exemplary embodiments, additional reverberation elements may be attached to the wall. The sound masking reverberation inducing element may be arranged in direct contact with a single wall or part of a wall, so that in certain exemplary embodiments the wall may act as a sound source. In certain exemplary embodiments, the sound masking reverberation inducing element may be disposed between walls in the wall assembly. The sound masking advantageously enables an increased noise/signal contrast, which results in a lower intelligibility of the speech perceived behind a single wall or part of a wall and a lower annoyance of the irritating sound.
In certain exemplary embodiments, a first set of features may be formed in and/or on the inner layer and a second set of features may be formed in and/or on the outer layer, for example, to block some annoying or distracting sounds and improve the "internal" acoustic effect. In certain exemplary embodiments, multiple sets of features can be formed in and/or on one or both layers of a two-layer wall assembly, where each set of features targets a different range to be eliminated and/or enhanced
Other natural characteristics of the wall assembly (including dimensions, space between adjacent upstanding walls, etc.) may also be selected to trigger the desired reverberation effect, for example, as described above.
As mentioned above, it should be understood that these more passive techniques may be used in addition to the active techniques discussed above, such as utilizing a single or double wall acoustic wall assembly.
The wall assembly can thus be made in the manner of an acoustic resonator with a specifically designed fundamental resonance frequency. As noted above, any suitable material may be used to construct the walls. For example, because glass is a naturally good resonator, certain exemplary embodiments can utilize a variety of resonant harmonics that are integer multiples of the fundamental frequency. Regardless of the material, adjusting the incoming sound via the features may help to interfere with the frequency range of speech and noise, so as to make it unintelligible and/or less annoying. For example, when processing speech or the like, those frequency ranges associated with consonants or voicelets may be targeted. Furthermore, because such wall assemblies are designed for selective acoustic interference, in certain exemplary embodiments, thin glass and more durable rigid joints may be used in the wall assembly. Advantageously, this configuration may make the overall design more robust and reliable. When glass is used, high tolerances may be desirable to help maximize the effectiveness of the acoustic resonance characteristics by avoiding leakage and the like.
The walls described herein may be partial walls, e.g., walls that leave open spaces between separated regions. That is, the acoustic wall and acoustic wall assembly may be full or partial height in different circumstances. Single or double sided siding may also be used. Further, while certain exemplary embodiments have been described in connection with walls and/or rooms, it should be understood that the techniques described herein may be used in connection with more general areas where there are no or fewer defined divisions or structurally defined divisions (e.g., in hospital wards where curtains separate two patient areas, in lobbies, between front and rear seats of a car, between different rows or areas of an airplane, etc.).
Although the assignee of the patent has used passive or active (e.g., computer-generated) reverberation to reduce the intelligibility of perceived speech, it has been found that further improvements are still possible. For example, the human brain is adapted to process reverberation by giving priority to early arriving signals. Furthermore, so-called phoneme recovery is known to help the brain to recover information of missing sounds or overlapping sounds. Both phenomena sometimes filter out the same time-delayed copies and preserve the intelligibility of the original speech signal. This in turn can hamper the effectiveness of simple reverberation. In the exemplary embodiments described below, another potentially more efficient method of disturbing intelligibility and reducing the annoyance of perceived speech is presented, which takes these issues into account.
Referring again to step S908 in fig. 9 and how intelligible interference frequencies may be generated, certain exemplary embodiments use a dynamic approach to masking signals applied over the original speech. The method uses one or a combination of any of the following methods: (1) constant time delay, (2) time-varying time delay (time-phase adjustment), (3) amplitude modulation, and (4) spectral filtering. The contribution of these effects can be tuned according to specific needs or requirements. For example, in an environment where a certain degree of silence and calm is desired (e.g., a hospital recovery room, etc.), the amplitude increase variation may be kept to a minimum, while in an area where a large amount of noise is expected (e.g., a hospital waiting room, a police department "trial room," etc.), the amplitude increase variation may be larger.
The above method has been found to produce reliable speech interference. However, sometimes a significant increase in perceived loudness of sound may occur, and the listener may feel annoyed by the increased loudness. It will therefore be appreciated that it would be desirable to further improve techniques for disturbing original speech without significantly increasing the loudness and potential annoyance of the original speech.
Humans tend to understand duplicate sounds (as long as they are similar in shape) as part of the original sound, effectively ignoring the informational content and focusing only on the increased loudness. This is called the precedence effect. However, the replica signal can be further modified to interfere with the informational content and help reduce the effects of the precedence effect. Certain exemplary embodiments thus improve the above-described techniques by selectively interference masking the speech signal. As will become more apparent below, such selective interference may occur in conjunction with phonemes, consonants, and/or other speech building blocks.
Certain exemplary embodiments use reverberation delayed oscillation frequencies in the several hertz range. This range is advantageous because it corresponds to the number of syllables per second in normal english speech. Thus, certain exemplary embodiments enable speech intelligibility to be substantially disturbed without adding a significant amount of noise. That is, it has been recognized that the information-bearing frequency of speech is in a different frequency range than the "disturbing" part, thus allowing speech content interference to occur for the former's goal, while the additional loudness caused by sound masking is at a low cost.
In some exemplary embodiments, the speech intelligibility interference mask signal may take the general pattern of the original speech signal. In certain example embodiments, the masking signal may be delayed relative to the original signal, and/or a plurality of pre-recorded voices may be added to the speech intelligibility interference signal (e.g., to form a perception of crowd noise). In certain exemplary embodiments, other sounds (such as the sounds described above and/or other natural sounds, sound "beautifiers," etc.) may be added to further improve the speech intelligibility interference effect.
FIG. 10 shows the phoneme frequencies of a single-voiced speech and a multi-voiced speech at the top and bottom portions thereof, respectively. It should be appreciated that in some exemplary embodiments, a lower graph may be added above the detected speech, e.g., to interfere with the understandability of the speech, etc.
Fig. 11 illustrates the voiceprint frequencies of different types of sounds, including different natural sounds and different sounds of speech, which may be added to the latter as a sound beautifier or the like, e.g., as described above.
In operation, a method for disturbing speech intelligibility includes receiving an original speech signal via a microphone or other listening device. The original speech signal includes a plurality of phoneme (building blocks of speech intelligibility) and has some basic level of intelligibility that can be perceived by a human listener. The raw speech signal is processed (e.g., using a hardware processor or other control circuitry) to identify frequency ranges associated with the phonemes that make up the raw speech signal. Various parameters may then be used to substantially alter the speech signal and form an intelligible interference masking signal. For example, the intelligibility-interfering signal may be generated to include intelligibility-interfering phoneme in the same frequency range as the phoneme constituting the original speech signal, and the level of intelligibility of the resulting perceived speech may be reduced by outputting the intelligibility-interfering signal including the generated intelligibility-interfering phoneme via the speaker. In some cases, intelligible interfering phoneme is generated within the frequency range of 0.02-8 Hz. In some cases, the intelligible interfering speech sounds are generated at a frequency of 2-6Hz (e.g., 4 Hz).
In some example embodiments, the intelligibility interference mask signal may be time delayed relative to the original speech signal, for example such that the intelligibility interference mask signal follows a general pattern of the original speech signal, is a time delayed copy of the original speech signal, is a time phase adjusted copy of the original signal, is an amplitude modulated version of the original speech signal, and so on. A constant time delay range of 0-150ms is preferred, 40-120ms is more preferred, and 60-110ms is more preferred. In some cases, an exemplary delay of 80ms may be optimal, and in other cases, an average delay of 80ms may be optimal. In certain exemplary embodiments, dynamic reverberation may additionally or alternatively be used, for example, such that the time delay oscillates over time.
In some exemplary embodiments, the gain relative to the original speech signal may additionally or alternatively be adjusted. Furthermore, the gain may also be modulated over time. For example, the intelligible interference mask signal may be generated such that the loudness of the intelligible interference signal oscillates over time. Preferably, the gain (corresponding to the modulated intelligible interfering signal added to the original speech signal) is not too large, since this may produce negative psychoacoustic effects, e.g. by producing too much loudness or interference. In some exemplary embodiments, the gain applied is up to double that of the corresponding original speech signal. In certain exemplary embodiments, the gain is or averages 0.05 to 0.25%, more preferably 0.10 to 0.20%, with an exemplary gain of 0.15%.
In certain example embodiments, the time delay and/or amplitude adjustment may be modulated at one or more given frequencies. For example, the time delay and/or amplitude adjustment may be modulated at or may average at an oscillation frequency of 1-10Hz, more preferably 2-6Hz, with 4Hz being an example. It should be appreciated that the modulation may be the same or different for time delay and amplitude adjustment in different exemplary embodiments. In certain example embodiments, the delay and/or amplitude modulation may be provided according to one or more algorithms. In certain example embodiments, the delay and/or amplitude modulation may be a result of gaussian modulation, random modulation, modulation according to a waveform (e.g., sine wave, square wave, etc.), step-wise modulation, modulation according to a predefined pattern (e.g., increasing then decreasing frequency oscillation, etc.), application of an algorithm, and/or the like. In certain exemplary embodiments, dynamic time delay modulation of 40-400Hz, more preferably 60-300Hz, and for example 80-230Hz may be used.
Certain exemplary embodiments may further comprise outputting, via the speaker, the additional masking sound signal and the intelligible interfering signal comprising the generated intelligible interfering phoneme. For example, the intelligible disturbance signal may be generated as a pre-recorded mix comprising a plurality of voices. In addition or alternatively, a sound beautifier or the like may be used.
In certain exemplary embodiments, this functionality may be incorporated into an electronic device. Fig. 12 is a block diagram of an electronic voice intelligibility interfering device according to some example embodiments. The electronic device may include or otherwise be coupled to a
As mentioned above, in some exemplary embodiments, other building blocks of speech may be targeted. For example, the fundamental frequency of speech is known to occur between 85Hz and 250 Hz. Above this low frequency "fundamental channel", there are additional speech building blocks that include (a) "inert" vowels that are primarily responsible for determining the energy phoneme of the voice "power", and (b) information-bearing consonants.
Consonants contain little energy, but are believed to be critical to intelligibility (at least when english and other languages are involved), for example in the form of phonological units that distinguish meaning, namely phonemes (defined by the position of both clarity and loudness) and frequency-dependent keys. In some cases, other speech building blocks (such as duration-related time bits) may also be targeted. Vowels occur between 350Hz and 2KHz and are primarily volume bearing blocks of speech. Targeting low-volume information to carry consonants and leaving high-volume vowels intact by means of spectral filters may further help to reduce annoyance during speech disturbances.
The various consonants differ in the degree of contraction of the acoustic cavity and the timing of articulation. Even so, most of them lie in the frequency range between 1.5kHz and 4 kHz. In this regard, FIG. 13 includes examples of the frequency dependence of various syllables, where each syllable includes a consonant and a vowel.
Although the initial phonetic conversions of key consonants differ for subsequent vowels, their phonetic understanding remains unchanged. This knowledge can be used to trigger speech interference based on the threshold frequency of the consonants, which can also be considered the primary information-bearing speech unit in some cases.
Thus, in certain exemplary embodiments, the generation of the masking signal may be triggered based on reaching a threshold frequency (e.g., about 1.5kHz) that is higher than the frequency of most vowels but lower than the frequency of most consonants. In certain exemplary embodiments, a preset frequency range of 1.2-2kHz may be effective in this regard. Such an approach may help prevent the duplication of most vowels that carry little information load but contribute to undesirable loudness, and may instead help focus the duplicate signal on information-bearing consonants. A high-pass acoustic filter, for example, may be used in this regard. The fig. 12 block diagram may be used in conjunction with such exemplary techniques, for example, provided that such a high pass acoustic filter is provided before the
In some exemplary embodiments, the masking signal may oscillate (time phased) in a manner to provide a delay between 20ms and 95ms, which corresponds to the Voice Onset Time (VOT) of most consonants. VOT is the time between the release of a "pause" consonant and the beginning of a vocalization. A modulation frequency for time phase adjustment in the range of 1-10Hz may be advantageous, 2-10Hz being more advantageous, 2-6Hz being even more advantageous, and of which 4Hz is one example believed to be optimal. In certain exemplary embodiments, amplitude modulation may also be implemented. Amplitude modulation of 10-100% of the original signal, and more preferably 40-90% of the original signal, has been found to be advantageous in this respect.
Some exemplary techniques that take into account internal reverberation will now be described. As described above, different rooms have potentially different acoustic characteristics, including potentially different T measured within the room60The value is obtained. At a high T60In a room of values, excessive reverberation can be a problem. For example, when high intelligibility of speech in a room is involved, glass is incorporatedA wall or window room may face even greater challenges: the internal reverberation from the highly acoustically reflective surface acts as a masking signal. Different rooms, including rooms with glass, have been found to have disturbing internal acoustic reverberation therein, especially in the low frequency range (e.g., 20-200 Hz). While there are some solutions available that can help address the disturbing reverberation in the interior room (including, for example, the use of various sound absorbing surfaces), these solutions tend to hinder the transparency of the glass and tend to add significant cost.
Additionally or alternatively, certain exemplary embodiments provide acoustic solutions for reducing (and sometimes even eliminating) disturbing acoustic reverberation caused by reverberation in the low frequency range in a room or region. For example, certain exemplary embodiments generate a copy of the original speech signal that has an equalized (or substantially equalized) loudness but no disturbing reverberation in the lower portion of the spectrum.
Fig. 14 is a block diagram of an electronic device that facilitates reducing disturbing reverberation in a room, according to some example embodiments. The electronic device may include or otherwise be coupled to a
This generates a modified version of the acoustic pattern corresponding to the original speech such that the level of the new combined sound is equal to or substantially equal to the combined level of the original sound and the disturbing reverberation. However, the undesirable reverberation is generally "cut out" of the resulting spectrum in the modified version of the acoustic mode, and thus there are no spikes therein.
It should be appreciated that the shape of the substantially cut-out signal may be square, sine wave pattern, gaussian pattern, etc. In certain exemplary embodiments, the shape of the substantially cut-out signal may be more precisely adjusted to match the shape of the reverberation waveform. In some cases, a single fundamental reverberation mode may be removed, while in other cases a wider frequency range will be removed. In certain exemplary embodiments, a delta function that causes abrupt ablation may be used in this regard.
Although fig. 14 shows the
Fig. 15 is a graph showing an exemplary mask signal (gray) superimposed on an original speech signal (black). Clones were recorded at an exemplary sampling rate of 8kHz (although other sampling rates may be used in other exemplary embodiments). It should be understood that fig. 15 shows only one example of how speech may be disturbed. That is, unless explicitly stated, time delays, amplitude modulations, etc. shown in and/or implied by the graph are provided by way of example.
A test chamber was constructed and certain exemplary techniques were evaluated. The test room is a typical gray board office with the HVAC fan temporarily disabled, the reverberation time is 0.4s, and no special sound insulation measures. The target speech signal is played using a Yamaha HS5 speaker positioned behind one of the walls with STC of 30. This signal was registered using a Crown Audio far-field microphone, processed using software, and played using the same speaker positioned 2 meters in front of the subject in the room. The software uses a combination of the following four audio effects: (1) constant time delay, (2) time-varying time delay (time-phase adjustment), (3) amplitude modulation, and (4) spectral filtering. The time delay, modulation frequency and modulation depth are all adjustable parameters. The speech stimulus is a block of 100 prerecorded, brief, 5-7 word long, unrelated, and syntactically and semantically correct utterances spoken by male speech at normal speech speeds. Utterances are presented individually to each of ten subjects who subjectively score the annoyance of perceiving speech recognition and masking sounds. All subjects were native english speakers with normal hearing. The following types of speech masking elements were used in the experiments: white Noise (WN), time delayed clones of the target speech signal (TD), a masking element that is an Optimized Combination (OC) of the four audio effects described above, and an OC masking element (OCB) that is augmented with a multi-speaker background.
In this test, the time delay of the OC mask element was set to 80 ms. The time delay phase adjustment and amplitude modulation are performed at a rate of 3 to 5 modulations per second. Pre-recorded speech from three speakers (two men and one woman) who are speaking simultaneously is used as the background for the OCB masking element. OC optimization is performed to change the clone signal to just enough basic cues to obscure the target speech, thereby making the target speech unintelligible with minimal additional levels of annoyance. The method is activated by voice and the strength of the masking signal is continuously self-adjusting according to the strength of the target speech.
The rate of 3-5 cycles per second of delay phase adjustment and amplitude modulation is similar to the number of syllables per second in normal english speech, which makes OC masking highly selective in interfering with the speech rhythm of the target speech, as described above. For comparison, and as described above, white noise and natural sounds are poor speech masking elements at moderate loudness, because their temporal patterns are different from those of normal speech. The use of spectral filters further minimizes the clutter associated with shadowing. The spectral filter balances the contributions of the spectral regions responsible for the high-energy vowels and the information-bearing consonants.
The scoring results are shown in fig. 16. For numerical ratings, the decibel level of all four masking elements is added to the decibel level of WN at which 50% of the sentences are perceived as unintelligible. With WN and TD masking elements, all ten subjects reported that bother and considerable cognitive fatigue were continuously noticed at the masking level when the speech was still audible but the words were not understood. In the case of OC and OCB masking, cognitive fatigue was not reported and the level of annoyance was greatly reduced. Most subjects stopped paying attention to the speech for the lack of content after masking for about 30s with OCB. Three subjects reported perceiving OC-masked speech as foreign language.
From the data of fig. 16, it should be appreciated that certain exemplary embodiments can provide a perceptually effective technique for speech masking, where cues related to speech intelligibility are obscured by temporal phase adjustment and amplitude modulation of the target signal. The relationship between perceived speech intelligibility and annoyance is evaluated in a subjective rating analysis. The method is advantageously activated by voice and automatically adjusted according to the psycholinguistic aspects of the speech and the acoustic phonetic cues. It may be used in a stand-alone sound shielding device, or as an integral part of an office wall in an architectural auditory space with low STC rating and high side wing losses, as well as in other applications discussed herein.
Methods of making the above and/or other walls and wall assemblies are also contemplated herein. For the exemplary active methods described herein, such methods may include, for example, building a wall, connecting a microphone and air pump to a sound masking circuit, and so forth. Configuration steps for the sound masking circuit are also contemplated (e.g., specifying one or more frequency ranges of interest, when/how to activate the air pump, etc.). The mounting operation may be used, for example, with respect to a microphone and/or air pump (including suspension of a speaker), etc. Integration with HVAC systems and the like is also contemplated.
Similarly, methods of retrofitting existing walls and/or wall assemblies are also contemplated and may include the same or similar steps. Retrofit installations are also contemplated herein.
Certain exemplary embodiments have been described in connection with an acoustic wall and an acoustic wall assembly. It should be appreciated that these acoustic walls and acoustic wall assemblies may be used in a variety of applications to alter perceived speech patterns, mask certain irritating sound components from adjacent areas, and so forth. Exemplary applications include, for example, acoustic walls and acoustic wall assemblies for rooms in a house; a room in an office; a limited waiting area for doctor's offices, airports, convenience stores, banks, shopping malls, etc.; an exterior acoustic wall and acoustic wall assembly for a home, office, and/or other structure; exterior elements of the vehicle (e.g., doors, sunroofs, etc.) and interior regions of the vehicle (e.g., such that a child seated in a rear seat may be acoustically masked while seated in a front seat, and vice versa); and so on. Sound shielding may be provided for noise from adjacent areas regardless of whether the adjacent areas are another room, outside the confines of the structure housing the acoustic wall and acoustic wall assembly, etc. Similarly, sound masking may be provided to prevent noise from entering adjacent regions of this or other classes.
In certain exemplary embodiments, a method for disturbing speech intelligibility is provided, the method comprising: receiving an original speech signal corresponding to an original speech via a microphone; generating an intelligibility interference masking signal containing the fuzzy cues of the original speech in the original speech signal; and reducing the level of intelligibility of the original speech signal by outputting an intelligibility interference masking signal comprising the blurred speech prompt via the loudspeaker.
In addition to the features of the previous paragraph, in some exemplary embodiments the intelligibility interference mask signal may be time delayed relative to the original speech signal, e.g., by 20-150ms, 80ms, etc.
In addition to the features of the previous paragraph, in certain exemplary embodiments, the time delay may oscillate over time, for example, the time delay oscillates over time relative to the original signal in the range of 80-230 ms.
In addition to the features of any of the preceding three paragraphs, in certain exemplary embodiments, the intelligible interference mask signal may be generated such that an amplitude of the intelligible interference mask signal is modulated over time.
In addition to the features of any of the preceding four paragraphs, in some exemplary embodiments, the intelligibility interference mask signal may be generated such that the gain corresponding to the intelligibility interference mask signal added to the original speech signal is 0.05-0.25%.
In addition to the features of any of the preceding five paragraphs, in certain exemplary embodiments, the time delay may oscillate at an oscillation frequency of 1-10Hz, for example, the time delay oscillates at an oscillation frequency of 2-6 Hz.
In addition to the features of any of the preceding six paragraphs, in certain exemplary embodiments, the ambiguous prompts may be generated at a frequency of 0.01-20Hz, such as at a frequency of 2-6 Hz.
In addition to the features of any of the preceding seven paragraphs, in certain exemplary embodiments, the method may further comprise outputting, via the speaker, the intelligibility interference mask signal and a pre-recorded mixture of the plurality of voices, e.g., the pre-recorded mixture of the plurality of voices including 2-7 different voices, 3 different voices, and so on.
In some exemplary embodiments, a speech intelligibility interfering device comprising control circuitry may be configured to implement the functionality of any of the preceding eight paragraphs.
In certain exemplary embodiments, a system may include the apparatus of the previous paragraph.
In certain exemplary embodiments, the wall may incorporate the system of the previous paragraph.
While the invention has been described in connection with what is presently considered to be the most practical and preferred embodiment, it is to be understood that the invention is not to be limited to the disclosed embodiment, but on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.
- 上一篇：一种医用注射器针头装配设备
- 下一篇：存储器电路