GMU:Procedural Cut/Assignment Object Recognition: Difference between revisions

From Medien Wiki
 
Line 17: Line 17:
* [https://github.com/cloud-annotations/training cloud annotations]
* [https://github.com/cloud-annotations/training cloud annotations]
* [[Tracking Motion Detection]] this page is very much outdated since machine learning has become a thing
* [[Tracking Motion Detection]] this page is very much outdated since machine learning has become a thing
=== Auto_detecting_words_in_srt ===
[[File:guichu.txt]]
Copy the txt file to a empty .py file :D (.py file can't be uploaded, I wonder why @_@)
#library srt link: https://github.com/cdown/srt
#pip install -U srt
#copy paste srt file as in the code (focus on the format at head and tail)
#it returns a sequence of subtitle classes object, contains index,time, content, etc.
#using for loop searching desired word in every object.content.
#it returns the time interval at the start and the end
#use srt.timedelta_to_srt_timestamp convert the time to normal one

Latest revision as of 22:59, 1 December 2019

This Assignment is a bit different. There is no precise instruction. Goal is to edit video according to image analysis, or more specifically: Object recognition. For this assignment everyone needs to communicate to coordinate the effort and share ideas how to reach that goal. Edit this page here.


Example

Object detection using OpenCV and YOLO.

Input

Output

File:Log dog.txt


Here is the good olde color tracking in the new fancy clothes of "heuristics" https://towardsdatascience.com/real-time-object-detection-without-machine-learning-5139b399ee7d