Feature-map prediction with autoencoder reconstruction is faster and uses less energy than direct frame prediction, but it is less accurate, and the efficiency claim rests on an unspecified baseline.
Sánchez Velázquez1, Mingbo Cai2, 3, Andrew Coney1, Álvaro J
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Benefits of Feature Extraction and Temporal Sequence Analysis for Video Frame Prediction: An Evaluation of Hybrid Deep Learning Models
Feature-map prediction with autoencoder reconstruction is faster and uses less energy than direct frame prediction, but it is less accurate, and the efficiency claim rests on an unspecified baseline.