philferriere/tfoptflow

Optical Flow Prediction with TensorFlow. Implements "PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume," by Deqing Sun et al. (CVPR 2018)

51
/ 100
Established

This project helps computer vision practitioners analyze motion in videos by estimating optical flow. It takes two consecutive video frames as input and produces a 2D vector map representing the motion of every pixel from the first frame to the second. This output is crucial for tasks like object tracking, recognizing actions, and segmenting objects in video footage.

530 stars. No commits in the last 6 months.

Use this if you need to precisely measure pixel-level motion between video frames for computer vision applications like surveillance, robotics, or video analysis, and require a robust, trainable, and portable solution.

Not ideal if you need an extremely optimized, frozen model for real-time inference in a production environment without further training or finetuning.

video-analysis computer-vision object-tracking motion-estimation action-recognition
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 25 / 25

How are scores calculated?

Stars

530

Forks

134

Language

Jupyter Notebook

License

MIT

Last pushed

Aug 01, 2019

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/philferriere/tfoptflow"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.