tesh – Speech

[Warning: mild profanity]

My program was based on the idea of a group voicechat like Skype or Discord, but for all those gamers without friends. It is a groupchat for one person, where all of your friends are computer synthesized voices! You just start up the program and run it while you play some competitive online game, and you can pretend you have friends watching and cheering you on or giving you a hard time if you do poorly. The program responds to your commentary while you are playing the game and react differently based on how you are doing, using your word choice to determine if things are going well or not. If you are doing well and talk about how nice things are going and how the other team is “getting rekt,” your fake friends will cheer you on! But on the flip side, if you start complaining about dying they will take notice of that and gently cyberbully you. If your mic quality isn’t great then your computer comrades can’t hear you and they make comments on that, and sometimes they just have some wisdom they need to bestow upon you that they were reminded of when you said some word. It is about simulating the chaos, interruption, and ridiculousness of a groupchat but now with no external friends required.

Ideally this program works best on a computer, running in the background while you do other things much like in a real voice chat or conference call. I found ways to modify the pitch and rates of speech of the synthesized voices to make them sound more diverse and feel like more people were part of the chat. I also loaded in different dialects of English as well as some non-English voices.

The program latches onto keywords for positivity, or “hyping,” negativity, or “mocking,” and the other 2 modes, mishearing and spouting non-sequiturs, is based off of any random word you say that does not meet the first 2 categories. To create the non-sequiturs I loaded in a text file full of one-liners (from this website) and then created Markov chains based off of those, finally subbing in a word from the newly created one-liner with something the user said to add relevancy.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
// The speech recognizer
var mySpeechRecognizer;
var mostRecentSpokenWord;
var mostRecentConfidence;
var mostRecentPhrase = [];
var lastTwoWords = ["1", "2"];
var responseArray = [];
var textSource = [];
var txt = "";
 
rm = new RiMarkov(4, true, false);
 
var xshift = 25;
var yshift = 20;
// The speech synthesizer
var myVoiceSynthesizer;
var myVoiceNames = ["Microsoft David Desktop - English (United States)", "Microsoft Zira Desktop - English (United States)", "Google US English", "Google UK English Female", "Google UK English Male", "Google Nederlands", "Google Deustch"];
 
// The RiTa Lexicon
var myRitaLexicon;
var bFoundRhymeWithMostRecentWord = true;
var bWord = true;
var aWordThatRhymesWithMostRecentWord = "";
var arrayOfRhymingWords = [];
var confidenceThresh = 0.3;
var mishearChance = 0.75;
var nonSeqChance = 0.7;
var responsePhrase = "";
 
//Prompts?
var mockArray = ["fuc", "fuk","f***", "f******", "s***", "sniped", "ouch", "oh", "impossible", "died"];
var hypeArray = ["nice", "sick", "sweet", "rad", "yes", "yeah"];
var hypeArray2 = ["I won", "we won", "we win", "sniped them", "did it", "get wrecked", "get rekt", "pretty good"];
 
//Responses
var mockResponseArray = ["you sure did it.", "lol, get rekt!", "scrub", "what, was that.", "delete your account", "time for a different game?", "I've never seen something so terrible."];
var hypeResponseArray = ["hell yes", "nice!", "sweet", "heck yeah", "woah!", "nice job!", "not the best I've seen but definitely not the worst", "pretty damn good!", "great job!", "kickass!"];
var nonseqArray = ["You know what they say,", "Oh that reminds me of how", "Like they always say,", "lol, like I always say,"];
var mishearArray = ["you do what with ", "you what your what?", "what did you just say to me?", "I think your microphones busted.", "you're cutting out!"];
 
//=========================================
function preload() {
  textSource = loadStrings('/addons/1liners.txt');
}
 
//=========================================
function setup() {
	createCanvas(520, 520);
	// Make the speech recognizer
 
	txt = join(textSource, '\n');
 
	rm.loadText(txt);
 
	mostRecentConfidence = 0;
	mostRecentSpokenWord = "";
	initializeMySpeechRecognizer(); 
 
	// Make the speech synthesizer
	myVoiceSynthesizer = new p5.Speech();
	myVoiceSynthesizer.setVoice(random[myVoiceNames]);
	myVoiceSynthesizer.interrupt=true;
 
	// Create the RiTa lexicon (for rhyming)
	myRitaLexicon = new RiLexicon();
}
 
//=========================================
function initializeMySpeechRecognizer(){
	mySpeechRecognizer = new p5.SpeechRec('en-US'); 
 
	mySpeechRecognizer.continuous = true; // do continuous recognition
	mySpeechRecognizer.interimResults = false; // allow partial recognition 
	mySpeechRecognizer.onResult = parseResult; // recognition callback
	mySpeechRecognizer.start(); // start engine
 
	console.log(mySpeechRecognizer);
}
 
//=========================================
function draw() {
	background(27);
 
	// Draw detected speech:
	fill(65);
	rect(xshift,yshift, width-(xshift*2),yshift+40);
	rect(xshift,yshift+70, width-(xshift*2),height-yshift*5.5);
 
	fill(180,215,240);
	textFont("Century Gothic");
	textSize(18);
	textAlign(LEFT);
	text("You entered the lobby..." , xshift+5, yshift*2);
	if(mySpeechRecognizer.resultString != undefined && mySpeechRecognizer.resultString.length <= 50){
		text("You said: \"" + mySpeechRecognizer.resultString + "\"", xshift+5, yshift*3+10); 
 	}
 	else{
 
		text("Say Something! ", xshift+5, yshift*3+10); 
 	}
	//findRhymeWithMostRecentWord();
 
	wordCheck();
	//print(hypeCheck());
	myVoiceSynthesizer.setRate(random(1.05,1.15));
	myVoiceSynthesizer.setPitch(random(0.5,1.45));
	myVoiceSynthesizer.setVoice(random(myVoiceNames));
 
	fill(51,139,213); 
	for(i=0;i<responseArray.length;i++){ text(responseArray[i], xshift+20, 110+(25*i)); } if(millis()%5000 >= 0 && millis()%5000 < 10){
		initializeMySpeechRecognizer();
	}
}
 
//=========================================
function keyPressed(){
	if (key === ' '){
		// Press the spacebar to reinitialize the recognizer.
		// This is helpful in case it freezes up for some reason.
		// If you have a lot of freezes, consider automating this.
		initializeMySpeechRecognizer();
	}
	if (key === 'M'){
		var sentences = rm.generateSentence();
		var rSentences = new RiString(sentences);
		var sentenceWords = rSentences.words();
		var sentencePos = rSentences.pos();
 
		var nonSeqWord = random(mostRecentPhrase);
		var wordSwap = new RiString(nonSeqWord);
		var pos = wordSwap.pos();
 
		for(i=0;i <sentenceWords.length; i++){ if (sentencePos[i] == pos[0]){ sentenceWords[i] = nonSeqWord; break; } } print(RiTa.untokenize(sentenceWords)); bWord = false; wordCheck(); myVoiceSynthesizer.setRate(random(1.05,1.15)); myVoiceSynthesizer.setPitch(random(0.5,1.45)); myVoiceSynthesizer.setVoice(random(myVoiceNames)); } } //========================================= function parseResult() { mostRecentConfidence = mySpeechRecognizer.resultConfidence; lastTwoWords[1] = "1"; lastTwoWords[2] = "2"; if (mostRecentConfidence > confidenceThresh){ // some confidence threshold...
		console.log (mySpeechRecognizer.resultString);
 
		// The Recognition system will often append words into phrases.
		// So the hack here is to only use the last word:
		mostRecentSpokenWord = mySpeechRecognizer.resultString.split(' ').pop();
 
		mostRecentPhrase = mySpeechRecognizer.resultString;
		mostRecentPhrase = RiTa.tokenize(mostRecentPhrase);
		print(mostRecentPhrase);
		print(mostRecentConfidence);
 
		if (mostRecentPhrase.length >= 2){
			lastTwoWords[1] = mostRecentPhrase[mostRecentPhrase.length-2];
			lastTwoWords[2] = mostRecentPhrase[mostRecentPhrase.length-1];
		}
		else if(mostRecentPhrase.length == 1){
			lastTwoWords[1] = "1";
			lastTwoWords[2] = mostRecentPhrase[mostRecentPhrase.length-1];
		}
		else{
			lastTwoWords[1] = "1";
			lastTwoWords[2] = "2";
		}
 
		bWord = false;
	}
}
 
//=========================================
 
function wordCheck(){
	if(bWord == false){
		responsePhrase = "";
 
		if(mostRecentConfidence > confidenceThresh && mostRecentConfidence < 0.7){ //Say statement from response array template responsePhrase = mishearArray[int(random(1,mishearArray.length-1))]; } else if(lastTwoWords[2] != "2"){ //check if should be hyping if(hypeCheck() == true){ hype(); } //check if should be mocking else if(mockCheck() == true){ mock(); } else if(random(1) > nonSeqChance){
				nonSeq();
			}
 
			else if(random(1) > mishearChance){
				mishearWord();
			}
		}
 
		bWord = true;
		if (responsePhrase != ""){
 
			updateTexts()
			myVoiceSynthesizer.speak(responsePhrase);
		}
	}
 
}
 
function mishearWord(){
	var similarWord="";
	var mishearNum = int(random(0,mishearArray.length));
 
	// if at least 1 word
 
	// generate similar word or keep original
	similarWord = lastTwoWords[2];
	if(random(1)<0.75){ similarWord = RiTa.similarBySound(lastTwoWords[2]); similarWord = similarWord[int(random(0,similarWord.length))]; } // Say word along with statement from array if(random(1) > 0.5){
		responsePhrase = ("What? "+similarWord+"??");
	}
	else{
		if (mishearNum > 2){
			responsePhrase = (similarWord+"?? "+mishearArray[mishearNum]);
		}
		else{
			responsePhrase = (mishearArray[mishearNum]+" "+similarWord+"??");
		}
	}
}
 
function hype(){
	var hypeRand = random(1);
	if(hypeRand > .6){
 
		responsePhrase = (random(hypeArray) +"! "+ random(hypeResponseArray)+"!");
	}
 
	else if(hypeRand > .3){
		responsePhrase = (random(hypeArray2) +"! "+ random(hypeResponseArray)+"!");
	}
	else{
		responsePhrase = random(hypeResponseArray);
	}
}
 
function hypeCheck(){
 
	if(lastTwoWords[1]!= "1"){
		for(i=0;i<hypeArray2.length;i++){
			if((lastTwoWords[1]+" "+lastTwoWords[2]) == hypeArray2[i]){
				print(lastTwoWords[1]+" "+lastTwoWords[2]);
				return true
			}
		}
	}
 
	for(i=0;i<hypeArray.length;i++){
		if(lastTwoWords[2] == hypeArray[i]){
			return true
		}
	}
 
	return false
 
}
 
function mock(){
 
	responsePhrase = (random(mockResponseArray));
 
}
 
function mockCheck(){
 
	for(i=0;i<mockArray.length;i++){
		if(lastTwoWords[2] == mockArray[i]){
			return true
		}
	}
 
	return false
}
 
function nonSeq() {
 
	var sentences = rm.generateSentence();
	var rSentences = new RiString(sentences);
	var sentenceWords = rSentences.words();
	var sentencePos = rSentences.pos();
 
	var nonSeqWord = random(mostRecentPhrase);
	var wordSwap = new RiString(nonSeqWord);
	var pos = wordSwap.pos();
	var newSent = "";
 
	for(i=0;i <sentenceWords.length; i++){ if (sentencePos[i] == pos[0]){ sentenceWords[i] = nonSeqWord; responsePhrase = (random(nonseqArray)+" \n"+RiTa.untokenize(sentenceWords)); break; } if(i==sentenceWords.length-1){ responsePhrase = ""; } } } function updateTexts(){ if(responseArray.length > 14){
		responseArray.shift()
	}
	responseArray.push(responsePhrase);
	if(responsePhrase.length>45){
		responseArray.push("");
		if(responseArray.length > 14){
			responseArray.shift()
		}
	}
 
}