models/yolov9/ #8484
Replies: 93 comments 208 replies
-
I am looking forward to using it in Ultralytics! |
Beta Was this translation helpful? Give feedback.
-
I am looking forward to using it in Ultralytics! |
Beta Was this translation helpful? Give feedback.
-
Are you planning to integrate only inference, or training and export too? |
Beta Was this translation helpful? Give feedback.
-
Can it be used for pose task? |
Beta Was this translation helpful? Give feedback.
-
What's the timeline look like for Yolov9 training pipeline? Excited to see further optimization on this project. |
Beta Was this translation helpful? Give feedback.
-
Intriguing improvements in the architecture. I'm quite excited to use this in my projects. |
Beta Was this translation helpful? Give feedback.
-
For video processing with yolov9, im using this function: |
Beta Was this translation helpful? Give feedback.
-
I am interested in knowing when yolo v9 might be expected to be available for image segmentation tasks. |
Beta Was this translation helpful? Give feedback.
-
Can this be used for multiclass classification ? |
Beta Was this translation helpful? Give feedback.
-
Can I train Yolov9 using p2.yaml configuration file? |
Beta Was this translation helpful? Give feedback.
-
Hi , i would like to know when you guys will integrate yolov9 with uralytics , thanks |
Beta Was this translation helpful? Give feedback.
-
Hello, I am training a custom model using Yolov9. I could not find any documentation about how to get the bounding box coordinates from the predictions. Could anybody help me with that? I am predicting using the detect.py script from the official Git repo. With Yolov8 I was able to use model.predict(), however if I use this on Yolov9 I get this error: TypeError: BaseModel.fuse() got an unexpected keyword argument 'verbose' Thank you in advance. |
Beta Was this translation helpful? Give feedback.
-
Hi , I'm really glad that yolo v9 has released, I have a pcb(circuit board) ,I use to detect defects in it.I have took image of it and annotated the defected areas , the defects is only 1 one means only one class , I have no idea how to train with only one image of that one PCB , all I want is to predict the defect in that particular image which I used for training , can you help me with this about how many epochs should I use or any other ways ? Please help I've been trying this for hours today , thank you |
Beta Was this translation helpful? Give feedback.
-
Also looking forward to the tfjs export |
Beta Was this translation helpful? Give feedback.
-
I'm currently planning to use YOLOv5 for a project, what would the benefits and drawbacks be of switching to YOLOv9? I'm partictularly interested in fps and speed |
Beta Was this translation helpful? Give feedback.
-
why yolov9s or yolov9m is not loading 😢? @pderrenger |
Beta Was this translation helpful? Give feedback.
-
please how to calculate and print accuracy from yolo model ? |
Beta Was this translation helpful? Give feedback.
-
Can I train YOLO v9 with Polygon Annotated Dataset ? |
Beta Was this translation helpful? Give feedback.
-
Is it mandatory to make all the training data 640x640 ? Because my data is larger than this resolution and always rectangular. |
Beta Was this translation helpful? Give feedback.
-
Hi, I got a problem!
Could anyone help me, plz |
Beta Was this translation helpful? Give feedback.
-
I am trying to export the YOLOv9e model to TFLite format using the code: model = YOLO('yolov9e.pt')
model.export(format='tflite') However, I get an error: ...
Depth of input (1408) is not a multiple of input depth of filter (256) for '{{node tf.nn.convolution_107/convolution}} = Conv2D[T=DT_FLOAT, data_format="NHWC", dilations=[1, 1, 1, 1], explicit_paddings=[], padding="SAME", strides=[1, 1, 1, 1], use_cudnn_on_gpu=true](Placeholder, tf.nn.convolution_107/convolution/filter)' with input shapes: [1,80,80,1408], [1,1,256,256].
...
ERROR: onnx_op_name: /model.22/cv1/conv/Conv
... When I look at the outputted ONNX model in Netron, the input shape for that node appears to be [1x256x80x80]. Am I missing something in the export? Exporting the YOLOv9c model to TFLite works without any issues. Any suggestions? |
Beta Was this translation helpful? Give feedback.
-
Hello :) Thanks a lot! |
Beta Was this translation helpful? Give feedback.
-
Hi there, I have some question about loading YoloV9 pretrained weight in Ultralytics. I have trained a yolov9 model (yolov9-e for detection problems) using code in the original repo of yolov9 at https://github.com/WongKinYiu/yolov9. Now, i want to load this weight using Ultralytics. But i have some error when loading it.
weight = "yolov9-e-best.pt"
model = YOLO(weight) I got the following error.
But it return the error too:
How can i fix this? |
Beta Was this translation helpful? Give feedback.
-
Hi, from wandb.integration.ultralytics import add_wandb_callback
import wandb
from ultralytics import YOLO
## Config
project = "YOLOv9"
run_name = "tseg_run0"
pretrain_weight = "yolov9e-seg.pt"
data_yaml = "seg_train.yaml"
epochs = 20
batch_size = 24
device = [0, 1, 2, 3]
workers = 16
## Init model
model = YOLO(pretrain_weight)
add_wandb_callback(model)
## Train
results = model.train(
data=data_yaml,
epochs=epochs,
imgsz=640,
batch=batch_size,
device=device,
workers=workers,
project=project,
name=run_name,
lr0=0.01,
lrf=0.01,
)
wandb.finish() When at epoch 10, I stop training using One more question, can I resume training using For example, in the previous training, I train with batch size 24, with 4 GPUs, but I Thank a lot! |
Beta Was this translation helpful? Give feedback.
-
Hi, I am using YOLOv9 model to detect text on images. I custom trained the model with very small custom dataset. The detections are not accurate as the dataset is very small(10 images). Are there any YOLO models pretrained to detect text on images. Or a annotated text on images dataset. |
Beta Was this translation helpful? Give feedback.
-
hi i got some error when export yolov9 custom model which is training by google colab to my local python script. Here is the error code. TypeError: ERROR ❌️ bgry_frisbee.pt appears to be an Ultralytics YOLOv5 model originally trained with https://github.com/ultralytics/yolov5. |
Beta Was this translation helpful? Give feedback.
-
Hi there! I got another issue when using the yolov9 model for my work. I integrate the yolov9 into my class as follows: from ultralytics import YOLO
class MyClass:
def __init__(
self,
weight_path,
device="cpu",
):
self.model = YOLO(weight_path)
self.device = device
def process(
self,
img_source: Union[str, np.ndarray, Image.Image],
conf: float = 0.25,
iou: float = 0.45,
crop_padding: int = 10,
) -> Dict:
infer_results = self._infer(img_source, conf=conf, iou=iou)
# work for final results
return final_results
def _infer(self, img_source, conf=0.25, iou=0.7):
results = self.model.predict(img_source, device=self.device, conf=conf, iou=iou)
return results[0] But when I call my class as follows, although I pass the device as "2", my model is always loading to card 0. Is anything wrong with my implementation? my_class = MyClass(weigth_path="path_to_weight", device="2")
my_class.process(...) |
Beta Was this translation helpful? Give feedback.
-
hi ,When I try to change the syntax I get this error from the export.py |
Beta Was this translation helpful? Give feedback.
-
I will custom train a yolo model with my first dataset like this Again I want to train that model with the second dataset is this a right approach, can i do like this? |
Beta Was this translation helpful? Give feedback.
-
WongKinYiu/yolov9@2169e2c |
Beta Was this translation helpful? Give feedback.
-
models/yolov9/
了解 YOLOv9,它是实时物体检测系统的最新成员,利用可编程梯度信息和 GELAN 架构实现了无与伦比的性能。
https://docs.ultralytics.com/models/yolov9/?h=yolov9
Beta Was this translation helpful? Give feedback.
All reactions