Modelpublic

egorchistov/deblurring-tsrn-rtvd

sourceHugging Faceupdated 4mo agoView on Hugging Face

0likes

Model Card

TSRN-RTVD: Real-Time Video Deblurring System

Install package:

shell

pip install git+https://github.com/msu-video-group/tsrn-rtvd

Then use it as you like:

python

import numpy as np
from tsrn_rtvd import TSRNRTVD

model = TSRNRTVD.from_pretrained("egorchistov/deblurring-tsrn-rtvd", device="cuda")

first = np.zeros((720, 1280, 3), dtype=np.uint8)
next_frame = np.zeros((720, 1280, 3), dtype=np.uint8)

model.reset(first)
result = model.process_next(next_frame)
deblurred_bgr = result.output_bgr