how can i play sound for each detection

[7assanx7]

`# Ultralytics YOLO 🚀, GPL-3.0 license

import torch

from ultralytics.yolo.engine.predictor import BasePredictor
from ultralytics.yolo.engine.results import Results
from ultralytics.yolo.utils import DEFAULT_CFG, ROOT, ops
from ultralytics.yolo.utils.plotting import Annotator, colors, save_one_box

class DetectionPredictor(BasePredictor):

def get_annotator(self, img):
    return Annotator(img, line_width=self.args.line_thickness, example=str(self.model.names))

def preprocess(self, img):
    img = (img if isinstance(img, torch.Tensor) else torch.from_numpy(img)).to(self.model.device)
    img = img.half() if self.model.fp16 else img.float()  # uint8 to fp16/32
    img /= 255  # 0 - 255 to 0.0 - 1.0
    return img

def postprocess(self, preds, img, orig_imgs):
    preds = ops.non_max_suppression(preds,
                                    self.args.conf,
                                    self.args.iou,
                                    agnostic=self.args.agnostic_nms,
                                    max_det=self.args.max_det,
                                    classes=self.args.classes)

    results = []
    for i, pred in enumerate(preds):
        orig_img = orig_imgs[i] if isinstance(orig_imgs, list) else orig_imgs
        if not isinstance(orig_imgs, torch.Tensor):
            pred[:, :4] = ops.scale_boxes(img.shape[2:], pred[:, :4], orig_img.shape)
        path, _, _, _, _ = self.batch
        img_path = path[i] if isinstance(path, list) else path
        results.append(Results(orig_img=orig_img, path=img_path, names=self.model.names, boxes=pred))
    return results

def write_results(self, idx, results, batch):
    p, im, im0 = batch
    log_string = ''
    if len(im.shape) == 3:
        im = im[None]  # expand for batch dim
    self.seen += 1
    imc = im0.copy() if self.args.save_crop else im0
    if self.source_type.webcam or self.source_type.from_img:  # batch_size >= 1
        log_string += f'{idx}: '
        frame = self.dataset.count
    else:
        frame = getattr(self.dataset, 'frame', 0)
    self.data_path = p
    self.txt_path = str(self.save_dir / 'labels' / p.stem) + ('' if self.dataset.mode == 'image' else f'_{frame}')
    log_string += '%gx%g ' % im.shape[2:]  # print string
    self.annotator = self.get_annotator(im0)

    det = results[idx].boxes  # TODO: make boxes inherit from tensors
    if len(det) == 0:
        return f'{log_string}(no detections), '
    for c in det.cls.unique():
        n = (det.cls == c).sum()  # detections per class
        log_string += f"{n} {self.model.names[int(c)]}{'s' * (n > 1)}, "

    # write
    for d in reversed(det):
        c, conf, id = int(d.cls), float(d.conf), None if d.id is None else int(d.id.item())
        if self.args.save_txt:  # Write to file
            line = (c, *d.xywhn.view(-1)) + (conf, ) * self.args.save_conf + (() if id is None else (id, ))
            with open(f'{self.txt_path}.txt', 'a') as f:
                f.write(('%g ' * len(line)).rstrip() % line + '\n')
        if self.args.save or self.args.show:  # Add bbox to image
            name = ('' if id is None else f'id:{id} ') + self.model.names[c]
            label = None if self.args.hide_labels else (name if self.args.hide_conf else f'{name} {conf:.2f}')
            self.annotator.box_label(d.xyxy.squeeze(), label, color=colors(c, True))
        if self.args.save_crop:
            save_one_box(d.xyxy,
                         imc,
                         file=self.save_dir / 'crops' / self.model.names[c] / f'{self.data_path.stem}.jpg',
                         BGR=True)

    return log_string

def predict(cfg=DEFAULT_CFG, use_python=False):
model = cfg.model or 'yolov8n.pt'
source = cfg.source if cfg.source is not None else ROOT / 'assets' if (ROOT / 'assets').exists()
else 'https://ultralytics.com/images/bus.jpg'

args = dict(model=model, source=source)
if use_python:
    from ultralytics import YOLO
    YOLO(model)(**args)
else:
    predictor = DetectionPredictor(overrides=args)
    predictor.predict_cli()

if name == 'main':
predict()`

how can i play sound for each detection or class in real time detection
i want to if class 0 detected play sound by mp3 file

[dh031200]

@7assanx7 Using the generator as shown below, you can use yolo frame by frame.
Set stram=True for generator.
Use r.boxes to get the data for each frame.

from ultralytics import YOLO

model = YOLO('yolov8n.pt')
results = model.predict(source=src, verbose=False, stream=True)
for r in results:
    if 0 in r.boxes.cls:    # boxes object contains bbox outputs.
        # YOUR CODE
        # play sound function

---

This is sample of r.boxes:

boxes: tensor([[ 1.2615e+03,  5.8268e+02,  1.3865e+03,  8.5076e+02,  1.0000e+00,  9.1142e-01,  1.5000e+01],
        [ 1.8397e+02,  7.3267e+02,  4.2540e+02,  8.5830e+02,  2.0000e+00,  8.9236e-01,  5.0000e+00],
        [-2.7735e-01,  7.4888e+02,  8.5042e+01,  9.0250e+02,  3.0000e+00,  7.8997e-01,  5.0000e+00],
        [ 6.6940e+02,  1.9053e+00,  7.5318e+02,  7.5838e+02,  4.0000e+00,  7.3303e-01,  1.7000e+01],
        [ 6.0532e+02,  5.6310e+02,  7.0006e+02,  8.0803e+02,  5.0000e+00,  7.5743e-01,  2.9000e+01]])
cls: tensor([15.,  5.,  5., 17., 29.])
conf: tensor([0.9114, 0.8924, 0.7900, 0.7330, 0.7574])
data: tensor([[ 1.2615e+03,  5.8268e+02,  1.3865e+03,  8.5076e+02,  1.0000e+00,  9.1142e-01,  1.5000e+01],
        [ 1.8397e+02,  7.3267e+02,  4.2540e+02,  8.5830e+02,  2.0000e+00,  8.9236e-01,  5.0000e+00],
        [-2.7735e-01,  7.4888e+02,  8.5042e+01,  9.0250e+02,  3.0000e+00,  7.8997e-01,  5.0000e+00],
        [ 6.6940e+02,  1.9053e+00,  7.5318e+02,  7.5838e+02,  4.0000e+00,  7.3303e-01,  1.7000e+01],
        [ 6.0532e+02,  5.6310e+02,  7.0006e+02,  8.0803e+02,  5.0000e+00,  7.5743e-01,  2.9000e+01]])
id: tensor([1., 2., 3., 4., 5.])
is_track: True
orig_shape: tensor([1080, 1920])
shape: torch.Size([5, 7])
xywh: tensor([[1324.0040,  716.7162,  125.0505,  268.0803],
        [ 304.6838,  795.4862,  241.4294,  125.6306],
        [  42.3822,  825.6887,   85.3192,  153.6231],
        [ 711.2910,  380.1426,   83.7773,  756.4745],
        [ 652.6879,  685.5687,   94.7351,  244.9300]])
xywhn: tensor([[0.6896, 0.6636, 0.0651, 0.2482],
        [0.1587, 0.7366, 0.1257, 0.1163],
        [0.0221, 0.7645, 0.0444, 0.1422],
        [0.3705, 0.3520, 0.0436, 0.7004],
        [0.3399, 0.6348, 0.0493, 0.2268]])
xyxy: tensor([[ 1.2615e+03,  5.8268e+02,  1.3865e+03,  8.5076e+02],
        [ 1.8397e+02,  7.3267e+02,  4.2540e+02,  8.5830e+02],
        [-2.7735e-01,  7.4888e+02,  8.5042e+01,  9.0250e+02],
        [ 6.6940e+02,  1.9053e+00,  7.5318e+02,  7.5838e+02],
        [ 6.0532e+02,  5.6310e+02,  7.0006e+02,  8.0803e+02]])
xyxyn: tensor([[ 6.5702e-01,  5.3951e-01,  7.2215e-01,  7.8774e-01],
        [ 9.5817e-02,  6.7840e-01,  2.2156e-01,  7.9472e-01],
        [-1.4445e-04,  6.9340e-01,  4.4293e-02,  8.3565e-01],
        [ 3.4865e-01,  1.7642e-03,  3.9228e-01,  7.0220e-01],
        [ 3.1527e-01,  5.2139e-01,  3.6461e-01,  7.4818e-01]])

[7assanx7]

thank you it's working but when i'm use gTTS library to speech in real time detection but detection very slow when detect the class that do this function ,do you have idea what is problem ?

[dh031200]

How about creating an audio file in advance and then playing that audio in a thread when an object is detected?
As far as I know gTTS takes quite a while to process. This way you don't need to use gTTS.

[7assanx7]

I used this method, but it plays the sound continuously without ending the content of the audio file like this :

from playsound import playsound
import threading

def play_audio():
playsound("welcome.wav")

for r in results:
if 0 in r.boxes.cls:
audio_thread = threading.Thread(target=play_audio)
audio_thread.start()

it work fine without lag in real time detection but as I said it play audio files many times without finishing the sound that make it like echo

[dh031200]

As far as I know pygame library can solve that problem. playsound cannot stop sound, but pygame can.
Unfortunately, I don't know the details, I've only used it a few times.

Install pygame:
pip install pygame

This is example:

from pygame import mixer

# Init mixer
mixer.init()

# Play mp3 file
mixer.Channel(0).play(mixer.Sound('sound.mp3'))

# Stop sound
if mixer.Channel(0).get_busy():
    mixer.Channel(0).stop()

# Stop and play
if mixer.Channel(0).get_busy():
    mixer.Channel(0).stop()
    mixer.Channel(0).play(mixer.Sound('sound.mp3'))

[glenn-jocher]

@dh031200 it sounds like you're experiencing overlapping audio because the new threads are being spawned for each detection without waiting for the previous audio to finish. To prevent this, you could implement a simple check to see if the previous sound is still playing before starting a new one.

One way to do this is to use a flag that indicates whether the sound is currently playing. You would set this flag to True when you start playing the sound and set it back to False when the sound finishes. You can modify the play_audio function to check this flag and only play the sound if it's not already playing.

Here's a conceptual example:

from playsound import playsound
import threading

# Flag to indicate if the sound is currently playing
is_playing = False

def play_audio():
    global is_playing
    if not is_playing:
        is_playing = True
        playsound("welcome.wav")
        is_playing = False

for r in results:
    if 0 in r.boxes.cls:
        # Start the audio thread if the sound is not already playing
        if not is_playing:
            audio_thread = threading.Thread(target=play_audio)
            audio_thread.start()

Please note that this is a simplified example and may not be thread-safe. In a multi-threaded environment, you should use proper synchronization mechanisms like locks to prevent race conditions. However, this should give you an idea of how you can prevent the sound from playing multiple times simultaneously.