猶記得一開始想借範例程式

 

測試麥克風︰

pi@raspberrypi:~/speech_recognition/examples $ python3 microphone_recognition.py 
ALSA lib pcm.c:2495:(snd_pcm_open_noupdate) Unknown PCM cards.pcm.front
...

Say something!

 

結果卡在『Say something!』？只好嘗試排錯矣！

Troubleshooting

Troubleshooting or dépanneuring is a form of problem solving, often applied to repair failed products or processes on a machine or a system. It is a logical, systematic search for the source of a problem in order to solve it, and make the product or process operational again. Troubleshooting is needed to identify the symptoms. Determining the most likely cause is a process of elimination—eliminating potential causes of a problem. Finally, troubleshooting requires confirmation that the solution restores the product or process to its working state.

In general, troubleshooting is the identification or diagnosis of “trouble” in the management flow of a corporation or a system caused by a failure of some kind. The problem is initially described as symptoms of malfunction, and troubleshooting is the process of determining and remedying the causes of these symptoms.

A system can be described in terms of its expected, desired or intended behavior (usually, for artificial systems, its purpose). Events or inputs to the system are expected to generate specific results or outputs. (For example, selecting the “print” option from various computer applications is intended to result in a hardcopy emerging from some specific device). Any unexpected or undesirable behavior is a symptom. Troubleshooting is the process of isolating the specific cause or causes of the symptom. Frequently the symptom is a failure of the product or process to produce any results. (Nothing was printed, for example). Corrective action can then be taken to prevent further failures of a similar kind.

The methods of forensic engineering are especially useful in tracing problems in products or processes, and a wide range of analytical techniques are available to determine the cause or causes of specific failures. Corrective action can then be taken to prevent further failure of a similar kind. Preventative action is possible using failure mode and effects (FMEA) and fault tree analysis (FTA) before full-scale production, and these methods can also be used for failure analysis.

 

後於文本中東察西找，發現了

Troubleshooting

The recognizer tries to recognize speech even when I’m not speaking, or after I’m done speaking.

Try increasing the recognizer_instance.energy_threshold property. This is basically how sensitive the recognizer is to when recognition should start. Higher values mean that it will be less sensitive, which is useful if you are in a loud room.

This value depends entirely on your microphone or audio data. There is no one-size-fits-all value, but good values typically range from 50 to 4000.

Also, check on your microphone volume settings. If it is too sensitive, the microphone may be picking up a lot of ambient noise. If it is too insensitive, the microphone may be rejecting speech as just noise.

The recognizer can’t recognize speech right after it starts listening for the first time.

The recognizer_instance.energy_threshold property is probably set to a value that is too high to start off with, and then being adjusted lower automatically by dynamic energy threshold adjustment. Before it is at a good level, the energy threshold is so high that speech is just considered ambient noise.

The solution is to decrease this threshold, or call recognizer_instance.adjust_for_ambient_noise beforehand, which will set the threshold to a good value automatically.

 

的建議？？且考之以程式庫參考文件︰

recognizer_instance.energy_threshold = 300 # type: float

Represents the energy level threshold for sounds. Values below this threshold are considered silence, and values above this threshold are considered speech. Can be changed.

This is adjusted automatically if dynamic thresholds are enabled (see recognizer_instance.dynamic_energy_threshold). A good starting value will generally allow the automatic adjustment to reach a good value faster.

This threshold is associated with the perceived loudness of the sound, but it is a nonlinear relationship. The actual energy threshold you will need depends on your microphone sensitivity or audio data. Typical values for a silent room are 0 to 100, and typical values for speaking are between 150 and 3500. Ambient (non-speaking) noise has a significant impact on what values will work best.

If you’re having trouble with the recognizer trying to recognize words even when you’re not speaking, try tweaking this to a higher value. If you’re having trouble with the recognizer not recognizing your words when you are speaking, try tweaking this to a lower value. For example, a sensitive microphone or microphones in louder rooms might have a ambient energy level of up to 4000:

import speech_recognition as sr
r = sr.Recognizer()
r.energy_threshold = 4000
# rest of your code goes here

 

The dynamic energy threshold setting can mitigate this by increasing or decreasing this automatically to account for ambient noise. However, this takes time to adjust, so it is still possible to get the false positive detections before the threshold settles into a good value.

To avoid this, use recognizer_instance.adjust_for_ambient_noise(source, duration = 1) to calibrate the level to a good value. Alternatively, simply set this property to a high value initially (4000 works well), so the threshold is always above ambient noise levels: over time, it will be automatically decreased to account for ambient noise levels.

 

方能作此初步驗證哩！！