@Robots-and-androids
Thanks Mark, I have this up and running! On a HP 4 core with 8 GB (no GPU), I get 1 frame per three seconds. --Thomas (the robot guy)
@drishtimamtani223
Running entirely on CPU 2018-06-12 10:28:47.598141: I T:\src\github\tensorflow\tensorflow\core\platform\cpu_feature_guard.cc:140] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 Finished in 34.79076075553894s warning: Error opening file (/build/opencv/modules/videoio/src/cap_ffmpeg_impl.hpp:834) After running the code, the video is not opening..Can you help with this
@Robots-and-androids
Mark, You may want to post a link to your "Complete Atom For Python Setup Tutorial!" I had to do a bit of searching to find it! Thanks, Thomas
@pallawirajendra
Thank you for the amazing video! Great learning.
@sgt.mcgragon359
Halo, What exactly happens when you give this command(9:09 - 9:30)?
N/A
Hello Mark, I was going through your videos, and was able to get it to work for the images, however, with video I simply couldn't. I tried both this script, and the next one (with webcam) and it works fine when I use it with the camera ( cv2.VideoCapture(0) ) but not with a video (.avi extension). When I run the script, the video starts playing, and it shows all boxes of different objects detected. It keeps going until it breaks due to the following error :
###
...
FPS 1.1
Traceback (most recent call last):
File "C:\Users\joni_\Desktop\dark\darkflow-master\thing.py", line 21, in <module>
results = tfnet.return_predict(frame)
File "C:\Users\joni_\Desktop\dark\darkflow-master\darkflow\net\flow.py", line 78, in return_predict
'Image is not a np.ndarray'
AssertionError: Image is not a np.ndarray
FPS 0.9
[Finished in 1819.183s]
###
I left the first FPS to show that it was running before. Any reason you can think of?
Thank you in advance, and great videos!
Edit: I should add that, with the first video of the series it worked!
@MuradAlQurishee
Thanks for the video. Marks I tried a millions of time to count multiple object of multiple categories. For example, in the tutorial video If I want to count number car and number traffic light separately, then how can I write my code. Please give me a demo code. I appreciate your cooperation. I apolozied for asking the code again and again.
@mvwej-0525
2018-02-01 22:01:58.378064: I C:\tf_jenkins\workspace\rel-win\M\windows-gpu\PY\35\tensorflow\core\common_runtime\gpu\gpu_device.cc:1105] Found device 0 with properties: name: GeForce GTX 1070 major: 6 minor: 1 memoryClockRate(GHz): 1.683 pciBusID: 0000:01:00.0 totalMemory: 8.00GiB freeMemory: 6.63GiB Alright, I am processing the video, which has resolution of 720P. I am running out of memory. Do you know how to make that work? Maybe having a lower resolution video?
@sitiida1961
Hi Mark, i watching your tutorial and i thing thats is very awesome, thaks for your serial yolo, Im not have a good skills in Python and I have question mark, i really hope you can help me, how if i want running this video process in web server and display it to web page like using flask maybe? Thankyou mark you are inspiring me 👍
@utkarshshukla7917
extremely useful stuff man thanks !!!!
@slavehaq7675
Hi Mark, I got an error when using atom for this tutorial. I am using tensorflow 14 using TF14 env in anaconda. Successfully run this tutorial in spyder under conda tf14 virtual env. But I still want to use Atom. How to add atom into the TF14 env?
the error:
[Command: python -u 'C:\Users\Kudak\Desktop\dark\darkflow-master\Tutorial part 3.py']
Traceback (most recent call last):
File "C:\Users\Kudak\Desktop\dark\darkflow-master\Tutorial part 3.py", line 13, in <module>
tfnet = TFNet(option)
File "C:\Users\Kudak\Desktop\dark\darkflow-master\darkflow\net\build.py", line 58, in _init_
darknet = Darknet(FLAGS)
File "C:\Users\Kudak\Desktop\dark\darkflow-master\darkflow\dark\darknet.py", line 13, in _init_
self.get_weight_src(FLAGS)
File "C:\Users\Kudak\Desktop\dark\darkflow-master\darkflow\dark\darknet.py", line 47, in get_weight_src
'{} not found'.format(FLAGS.load)
AssertionError: bin/yolo.weights not found
@shivangmathur8776
Great Video. Is there a way to save the output video to my local system?
@Orbis92
Thanks for this great tutorial so far.... I wanted to leave a comment on the last part, but I just saw, there are still 4 more to watch :) I probably missed it, but what GPU you are using?
@fahadahmed1706
Sir my video is not displaying in window as yours did in the last,i tried to run it again but it finishes after 75 seconds kindly guide me code works fine but cant get video displayed.
@drvlog4285
hey mark. It has been a great help up to now. BTW I was trying to do this in Wing IDE instead in atom. how can I get the video processed then. Thanks in advance
@eeshaanjain696
Hey, my code compiled but the video didn't open. It showed Finished in 12.xy seconds but nothing after that Any fix
@shivamsahil3660
Hello mark! Even though YOLO is the fastest object detection algorithm, I try to run it it shows 0.5 FPS which is very very slow... another algorithm using tensorflow works fine with my laptop but I can't figure out what's wrong with my yollo I used the same code you did... just changing capture('from_pre_existing_video_file') to real time and rest same.
@Narutorid1
Hi mark, I have a 8gb ram and gtx950m graphic card on my laptop i tried to do the code and it runs without giving errors but the video doesn't pops out. Any idea why? I'd really appreciate if you reply to me as I am doing similar yolo fall detection for my final year project your videos are seriously useful
@dudeking1000
is there a way to only output bounding boxes with a specific label (say cars, I ask this because the model is already trained to detect cars) or do we need to train a completely different model from scratch to detect cars? Thank you for the video
@aa-xn5hc