SCALED surrogate gradient is reinterpreted as a projection-based first-order local approximation of non-differentiable video codecs, enabling effective training of full neural wrappers with BD-Rate gains up to 23.59% on x264.
Efficient evaluation of quantization-effects in neural codecs
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
CATRF inserts standard codecs into the training loop of triplane radiance fields via straight-through estimation so the features adapt to codec distortions and achieve better rate-distortion performance for volumetric streaming.
citing papers explorer
-
A Projection-Based Surrogate Gradient Interpretation for Neural Codec Wrappers
SCALED surrogate gradient is reinterpreted as a projection-based first-order local approximation of non-differentiable video codecs, enabling effective training of full neural wrappers with BD-Rate gains up to 23.59% on x264.
-
CATRF: Codec-Adaptive TriPlane Radiance Fields for Volumetric Content Delivery
CATRF inserts standard codecs into the training loop of triplane radiance fields via straight-through estimation so the features adapt to codec distortions and achieve better rate-distortion performance for volumetric streaming.