AdaptiveFaceTracking

Computer vision project attempting to discover and track faces in unconstrained video. https://sites.google.com/vt.edu/adaptivefacetracking/home

As of right now face detection is the limiting factor

###Training Tf Models

cd tf/tensorflow/
source settings.sh
./run.sh

Check run.sh script for settings used.

###Running Face Tracking Make sure to place a relevant BigBangTheory clip in the root repo directory in .mp4 format. Set the video_path variable in detect_and_label_video.py

video_path = "./TBBT_S10E16.mp4"

Then run the script

./run_video.sh

Name		Name	Last commit message	Last commit date
Latest commit History 54 Commits
.idea		.idea
audio		audio
data		data
face-alignment		face-alignment
faces_in_frames		faces_in_frames
frames		frames
labeled_bbt		labeled_bbt
output_frames		output_frames
pickled_boxes		pickled_boxes
raw_bbt		raw_bbt
tf		tf
video_out		video_out
.gitignore		.gitignore
README.md		README.md
audio_classifier.py		audio_classifier.py
box_in_frame.pick		box_in_frame.pick
box_resize.py		box_resize.py
cnn_class.py		cnn_class.py
cnn_class.pyc		cnn_class.pyc
detect_and_label.py		detect_and_label.py
detect_and_label_video.py		detect_and_label_video.py
get_video_from_img.sh		get_video_from_img.sh
person_in_frame_box.pick		person_in_frame_box.pick
run_detect_and_label.sh		run_detect_and_label.sh
run_video.sh		run_video.sh
track_pickled_boxes.py		track_pickled_boxes.py
tracked_next_frame_box.pick		tracked_next_frame_box.pick
video_to_img.py		video_to_img.py

gavincangan/AdaptiveFaceTracking

Folders and files

Latest commit

History

Repository files navigation

AdaptiveFaceTracking

About

Topics

Resources

Stars

Watchers

Forks

Languages