drewnf / tensorflow_object_tracking_video Goto Github PK
View Code? Open in Web Editor NEWObject Tracking in Tensorflow ( Localization Detection Classification ) developed to partecipate to ImageNET VID competition
License: MIT License
Object Tracking in Tensorflow ( Localization Detection Classification ) developed to partecipate to ImageNET VID competition
License: MIT License
In Utils_Video.py line 5:
import utils_image should be changed to: import Utils_image.
Dear Ferri:
Thank you for your codes.
I am using Tensorflow with opencv3.3.1, and I changed some of the codes according to the demandas of different opencv versions (for example, cv2.cv.CV_CAP_PROP_FRAME_COUNT is changed to cv2.CV_CAP_PROP_FRAME_COUNT). After these minor changes, the code can be run without errors in my computer.
However, the output seems a little strange when I am running VID_yolo.py (the 3rd step in your README.md). The results are as follows:
Starting Loading Results
[==========================================================] 100% Time: 0:00:00
Finished Loading Results
Computing Final Mean Reasults..
Class:
nan
Max Value:
nan
Min Value:
nan
Elapsed Time:6 Seconds
Running Completed with Success!!!
It is a little confusing why it gives out NANs for I haven't changed any parameters in your code.
Also, although I saw the frames output, it is not quite the same with that in the folder /video_result, and the green bounding box is not seen.
Could you kindly tell me how to solve this problem? many thanks.
Hello, DrewNF
Any python 3 version will be appreciated :)
hey.
your code can't run on python3.
it have some code style error .
eg. python 2 use print
but python 3 use print()
when I run program, Terminal has a message about no kaffe module. waht should i do?
The first run I made with:
python VID_yolo.py --path_video video.mp4
I got this error:
Traceback (most recent call last):
File "VID_yolo.py", line 11, in
import Utils_Video
File "/home/roberto/Programación/TENSORFLOW/Tensorflow_Object_Tracking_Video-master/Utils_Video.py", line 5, in
import utils_image
ImportError: No module named utils_image
I changed this
import utils_image
To this
import Utils_Image
In Utils_Video.py
And now I get this error:
Traceback (most recent call last):
File "VID_yolo.py", line 11, in
import Utils_Video
File "/home/roberto/Programación/TENSORFLOW/Tensorflow_Object_Tracking_Video-master/Utils_Video.py", line 5, in
import Utils_Image
File "/home/roberto/Programación/TENSORFLOW/Tensorflow_Object_Tracking_Video-master/Utils_Image.py", line 10, in
import matplotlib.pyplot as plt
File "/home/roberto/.local/lib/python2.7/site-packages/matplotlib/pyplot.py", line 72, in
from matplotlib.backends import pylab_setup
File "/home/roberto/.local/lib/python2.7/site-packages/matplotlib/backends/init.py", line 14, in
line for line in traceback.format_stack()
File "/home/roberto/.local/lib/python2.7/site-packages/matplotlib/backends/init.py", line 16, in
if not line.startswith(' File "<frozen importlib._bootstrap'))
UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 32: ordinal not in range(128)
What's wrong? Thanks in advance!!!
hello, firstly, I find that your code have import error
. The reason should be that the names are not uniform between Utils_Video.py
and utils_video.py
or Utils_Image.py and utils_image.py
. Please check it.
Importantly, I run python VID_yolo.py --path_video xxx.mp4
, but it shows error:
Opening File Video:2.mp4
could Not Open : 2.mp4
Traceback (most recent call last):
File "VID_yolo.py", line 120, in <module>
main()
File "VID_yolo.py", line 108, in main
frame_list, frames = utils_video.extract_frames(args.path_video, args.perc)
TypeError: 'NoneType' object is not iterable
why can not open a mp4 file? can you give some advises?
The google Drive link to the YOLO weight file is broken (Section 3i). Could you update the link? Thanks!
How to fine tune your model?
I don't have sufficient data to retrain your model from scratch.
I want to fine tune your model on my data which has only two classes ?
Have you changed the basic TENSORBOX to run on multiclass? If you did, do you have some documentation on how to use your updated version of TENSORBOX with multiclass? If you didn't, on which network do you base your multiclass detection, is it YOLO? And what was the purpose of using TENSORBOX?
Thank you...
Hello,
I'm testing your code for videos and when I executed the script "train_multiclass.py" to obtain the model, it gave me the following error:
Traceback (most recent call last):
File "TENSORBOX/train_multiclass.py", line 610, in <module>
main()
File "TENSORBOX/train_multiclass.py", line 607, in main
train(H, test_images=[])
File "TENSORBOX/train_multiclass.py", line 555, in train
test_output_to_log = train_utils_multiclass.add_rectangles(H,
AttributeError: 'module' object has no attribute 'add_rectangles'
Is this expected?
Probably we should use "train_utils.add_rectangles" I guess?
Thanks in advance.
Dear Ferri:
Thank you for your codes. However, I found that the link to Inception has been invalidated. If I can, can I ask you to upload this link? many thanks.
It seem there is an indentation mistake when you construct graph
reregress is an option of rezoom
if H['use_rezoom']:
pred_boxes, pred_logits, pred_confidences, pred_confs_deltas, pred_boxes_deltas = build_forward(H, tf.expand_dims(x_in, 0), googlenet, 'test', reuse=None)
grid_area = H['grid_height'] * H['grid_width']
pred_confidences = tf.reshape(tf.nn.softmax(tf.reshape(pred_confs_deltas, [grid_area * H['rnn_len'], H['num_classes']])), [grid_area, H['rnn_len'], H['num_classes']], name='pred_confidences')
pred_logits = tf.reshape(tf.nn.softmax(tf.reshape(pred_logits, [grid_area * H['rnn_len'], H['num_classes']])), [grid_area, H['rnn_len'], H['num_classes']])
if H['reregress']:
pred_boxes = pred_boxes + pred_boxes_deltas
else:
pred_boxes, pred_logits, pred_confidences = build_forward(H, tf.expand_dims(x_in, 0), googlenet, 'test', reuse=None)
must be
if H['use_rezoom']:
pred_boxes, pred_logits, pred_confidences, pred_confs_deltas, pred_boxes_deltas = build_forward(H, tf.expand_dims(x_in, 0), googlenet, 'test', reuse=None)
grid_area = H['grid_height'] * H['grid_width']
pred_confidences = tf.reshape(tf.nn.softmax(tf.reshape(pred_confs_deltas, [grid_area * H['rnn_len'], H['num_classes']])), [grid_area, H['rnn_len'], H['num_classes']], name='pred_confidences')
pred_logits = tf.reshape(tf.nn.softmax(tf.reshape(pred_logits, [grid_area * H['rnn_len'], H['num_classes']])), [grid_area, H['rnn_len'], H['num_classes']])
if H['reregress']:
pred_boxes = pred_boxes + pred_boxes_deltas
else:
pred_boxes, pred_logits, pred_confidences = build_forward(H, tf.expand_dims(x_in, 0), googlenet, 'test', reuse=None)
Hi , you mentioned using temporal information in the future. That's a nice idea.
I found this project related. Do you think it's possible to utilize in tensorbox?
https://github.com/Guanghan/ROLO
Good luck with your thesis.
This is the worst code I have ever seen, no one
How can i get track back if it loses target
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.