Skip to content

sumitb/YouTube2Action

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 

Repository files navigation

YouTube2Action

======= Deep Learning object detectors in YouTube clips to identify action in the form of Subject-verb-Object triplets.

Dependency

Dataset

Instructions

  • YouTube2Action contains latest release.
  • Detailed code structure
    • data: This directory act as a place to store intermediate results and final output.
    • src
      • U2_data.py: Python script to obtain images from YouTube links using ffmpeg, youtube-dl

      • detect_all.m: MATLAB script to run LSDA

      • sub_obj.m: MATLAB script to obtain Subject-Objectsfrom data/subject_object.mat

      • verb_Extractor.sh: Bash script to run LRCN

      • svo_triplet.m: MATLAB script to combine Subject-Object and correpsonding verbs.

      • word.py: Python script to run evaluation and obtain final_svo.txt

      • classify_video.py, detect10k_demo.m: Libraries code which we configured for our use.

Citing

"LSDA: Large Scale Detection Through Adaptation." J. Hoffman, S. Guadarrama, E. Tzeng, R. Hu, J. Donahue, R. Girshick, T. Darrell, and K. Saenko. Neural Information Processing Systems (NIPS), 2014. "Long-term recurrent convolutional networks for visual recognition and description." Donahue, J., Hendricks, L. A., Guadarrama, S., Rohrbach, M., Venugopalan, S., Saenko, K., & Darrell, T. (2014).

About

Deep Learning action detectors in videos

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published