Official YOLO v7 Pose Estimation | Windows & Linux

[@kenzaba]

thanks a lot, i've tried this code on colab and i'm getting this type of error: ( File "detect.py", line 117, in detect s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to string IndexError: list index out of range). could help me please?!

[@kapilsaxena1]

Great video, can this be done on mouse or rat behaviour videos. We have several of those videos where we want to detect of mouse is moving sleeping/resting or freezing?

[@TheCodingBug]

Yes. You can write a program for that but you'll have to train custom pose estimation model on that.

[@boppadapusaiteja8074]

Hat's off great I like the way you are demonstrate the codes.👍👍

[@jonathantobias4244]

Good video. How do you create a custom dataset for pose estimation in YOLO format?

[@gadgetboyplaysmc]

Would love to know as well! please!

[@lordfarquad1530]

Thank you for video ,looking forward for the tracking an object video , How do we train the official yolov7 instance segmentation on custom ? or how to even test yolov7-mask on webcam

[@TheCodingBug]

YOLOv7 instance segmentation video is scheduled for Monday.

[@lordfarquad1530]

@@TheCodingBug hello , is there any update regarding the video? Thank you so much

[@TheCodingBug]

@@lordfarquad1530 It will be uploaded by the end of this month.

[@hughmungus8784]

Very cool my friend!

[@akremgomri9085]

No errors for me so far, but when I test with images or a video, no one is detected inside the frame. Weird !

[@sezermezgil9304]

Hi great tutorial and I have one question. How do I run it in real time?

[@TheCodingBug]

Instead of video file name as source, use 0 for webcam along with --view-img

[@abdelhamidazanzal4403]

thanks a lot for this great tutorial really like it! I have a question about evaluation of the model how can we evaluate the model and make sure that it will suit on a real time?

[@TheCodingBug]

I don't understand what you mean by evaluation? If you are looking for metrics on validation split, its available in the official paper. If you are asking for speed for real-time, well its almost real-time (with FPS 12-15).

[@abdelhamidazanzal4403]

@@TheCodingBug thanks for the answer. Yes I'm looking for the metrics ( confusion matrix..) I will check the official paper

[@mihaelaplats]

Cool video, could you please help on how to use HRNet for detecting pose on custom dataset??

[@TheCodingBug]

I don't understand. Why would we need pose estimation on custom dataset? Is it for some specific use case?

[@mihaelaplats]

@@TheCodingBug yes, to use it with photos taken with webcam

[@mihaelaplats]

Same usage with "detect" of yolo and any random picture

[@TheCodingBug]

@@mihaelaplats same code would work for that!!! You do not need to retrain the model.

[@mihaelaplats]

@@TheCodingBug yesss i know, but i don't know how to call hrnet with any picturw

[@sutranvan9203]

I am calculating the object's velocity. Is there any way to increase velocity accuracy? (eg increase fps)

[@faatihahrs]

Is this can run in google colab? I've been trying to use it on colab but always get this error Traceback (most recent call last): File "detect.py", line 195, in detect() File "detect.py", line 116, in detect s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to string IndexError: list index out of range

[@kenzaba]

hello i've tried the same thing, i get this error: File "detect.py", line 117, in detect s += f"{n} {names[int(c)]}{'s' * (n > 1)}, " # add to string IndexError: list index out of range do you know a solution for that ??

[@LordWildbeast]

I met the same problem, and I don't know the solution yet

[@sakinamosavi1104]

I faced the same error, Could you solve it. If yes, please share the solution.

[@muhammadraza1374]

Can we detect multiple objects & pose estimation for only person in this repo ?

[@thanossiouras1135]

Nice video! could you please tell me how to extract the key point per joint in a file? thanks

[@gajanen9734]

Hi, Can we do this in Python in Visual Studio editor?

[@TheCodingBug]

Yes

[@alaaali5739]

I try it on webcam but I have a low fps..how can I fix it ?!

[@benaamediajo]

run on GPU not cpu, by adding --device 0 if available

[@abrarluvrabit]

another great video thanks alot Sir for helping us (Students) in the object detection field. Can you please also make a video on a Multi-label object dection in a single bounding box. I have a dataset that involves images with bounding boxes which have more than 1 label. For example, I have 5 classes (0-4), then an image might have a bounding box with label "2" and another might have a bounding box with two labels "2, 4" ie it belongs to both categories. so how can we show two labels or classes on a single bounding box? please make a video on this or atleast provide some solution in your ans here i will be very thankful to you.

[@TheCodingBug]

You can add two bounding boxes on same object with different labels or define single multi-word label and use that for a bounding box.

[@moahaimen]

is it possible to detect certain pose? and make alarm on it

[@TheCodingBug]

Yes it is.

[@moahaimen]

@@TheCodingBug how ? can i get you help please? i need to do that

[@akremgomri9085]

It doesn't seem to be working for me

[@brunospfc8511]

Thanks you sir, if you could, please do the instance on this Yolov7, thanks

[@TheCodingBug]

It's scheduled for the next week.

[@LordWildbeast]

Sir, how to run it from colab?

[@nemiw4429]

Please tell your Indian countrymen that's how you do your job as a KZbin video producer and publisher. From the title it in English -not half Indian and half English - production quality and detail. Job well done. I dont want to appear rude but it has become, as you can imagine, frustrating when you click on an english title, except a quality video. What you get is poor quality (which is not the main issue at all, everybody has to start), but half the text spoken is in English and half in Indian.

[@TheCodingBug]

I hope things would get better :) PS: I am not Indian.

[@Nostalgia-futuro]

Do You have linkedin or a way to contact you