Leveraging Computer Vision to Encode Time-Series Data as Images for Visual Recognition Algorithms

แชร์
ฝัง
  • เผยแพร่เมื่อ 1 ธ.ค. 2024

ความคิดเห็น • 9

  • @hannesloots5075
    @hannesloots5075 2 ปีที่แล้ว +1

    Great summary of the work you have done. I am also thinking of using all three models in a image classification model but instead of having 3 seperate channels using them to represent three colours in the image thereby only having to using one image to process. Adding additional data about the time series in seperate channels sounds like a great idea. I will definitely try that as well. Thank you again for sharing your work 🙏

  • @ethanhunt8989
    @ethanhunt8989 3 ปีที่แล้ว +1

    Can you provide a link for the code implementation of GAF matrix used with CNN to forecast time series data.

  • @TheZeroomg
    @TheZeroomg 2 ปีที่แล้ว +1

    Isn't the space complexity transformed from 1D to 2D a downside?

    • @MatloobKhushi
      @MatloobKhushi  2 ปีที่แล้ว

      How please explain? 2D world makes more sense in many domains such as imaging.

    • @TheZeroomg
      @TheZeroomg 2 ปีที่แล้ว

      @@MatloobKhushi Hello, i'm just curious if the time cost and space cost of GAF pre-processing an issue in some low-latency application?

  • @ichrakkallala8705
    @ichrakkallala8705 2 ปีที่แล้ว

    Hello! is there a python code you share? I want to know how to concatenate into a multi-channel image and details about the CNN model, please.
    Best regards

    • @MatloobKhushi
      @MatloobKhushi  2 ปีที่แล้ว

      I think we haven't published the code

    • @randomdude79404
      @randomdude79404 2 ปีที่แล้ว

      @@MatloobKhushi Could you publish it ?