Skip to content

Releases: naseemap47/CustomActionRecognition-TensorFlow-CNN-LSTM

v1.1

22 Feb 05:12
@naseemap47 naseemap47
v1.1
b3fee5c
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.
Compare
Choose a tag to compare
v1.1 Latest
Latest

v1.1 🚀

  • Save image size, seq length and class names to model.h5 meta data
  • Removed image size, seq length and class names from inference.py
    • It will load from model.h5 meta data
Assets 2
Loading

VideoGenerator

10 Feb 12:06
@naseemap47 naseemap47
v1.0
fa1aa5f
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.
Compare
Choose a tag to compare
VideoGenerator

Added VideoGenerator

  • Why we need VideoGenerator 🤔?
    • Previously we are unsing Data Extraction technic, take all data and stored in array.
    • But if we have large data, we will get RAM outoff memory Error.
    • Data Extraction take 60-80% of total time.
  • Advantages of VideoGenerator 🥳
    • Videogenerator solve all of these problems.
    • Its same like Image-generator
    • Option to add Data Augmentation
Assets 2
Loading