Activity Recognition from Egocentric Photo-Streams

Project to recognize human activities from egocentric perspective.
Pre-trained models are exploited for the recognition.

Check README.md located in archive for more details.

Pre-trained Models

Download the folder with the pre-trained models from this link.

Please refer to requirements.txt.

Cell splits the input video located in ./input/ to frames stored ./input/frame/.
Cell read frame by frame located in ./input/frame/, predict labels, put predicted labels in the dedicated frame as text.
Cell read the recognized frames located in ./input/frame/ and build video stored in ./output.

Unable to find the correct tensorflow-gpu and keras versions.
Some pre-trained models are not well trained.
Maybe retrain the models.

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
archive		archive
brightside		brightside
datasets		datasets
evaluation		evaluation
experiments		experiments
frames		frames
ntcir		ntcir
testing		testing
training		training
utils		utils
23122021.ipynb		23122021.ipynb
24122021.ipynb		24122021.ipynb
25122021.ipynb		25122021.ipynb
Crossvalidation Results.ipynb		Crossvalidation Results.ipynb
Dataset split.ipynb		Dataset split.ipynb
Demo.ipynb		Demo.ipynb
NTCIR Dataset Exploration.ipynb		NTCIR Dataset Exploration.ipynb
README.md		README.md
requirements.txt		requirements.txt