A Deep Learning post-processor with a perceptual loss function for video compression artefact removal
Dataset available: Request
The "models.py" file contains: Spatial Artefact Reduction Model, Temporal Artefact Reduction Model and ProxyVQA Model. These models can be instanciated by:
from models import SpatialSuppression, TemporalSuppression, VideoQualityAssessment
The savedModels.py folder shows how these models can be loaded with the weights presented in the literature.