What AI says about All-In-One-Deflicker

The project implements blind video deflickering using neural filtering with a flawed atlas, as described in CVPR 2023. It ships code, datasets, pretrained weights, and step-by-step instructions for training and inference, including support for segmentation masks and an environment setup (Python 3.10, PyTorch 1.12, CUDA). The goal is to produce flicker-free video outputs by removing temporal flicker artifacts from input videos.