TY  - CONF
AU  - Fei Yang
AU  - Luis Herranz
AU  - Yongmei Cheng
AU  - Mikhail Mozerov
A2  - CVPR
PY  - 2021//
TI  - Slimmable compressive autoencoders for practical neural image compression
BT  - 34th IEEE Conference on Computer Vision and Pattern Recognition
SP  - 4996
EP  - 5005
N2  - Neural image compression leverages deep neural networks to outperform traditional image codecs in rate-distortion performance. However, the resulting models are also heavy, computationally demanding and generally optimized for a single rate, limiting their practical use. Focusing on practical image compression, we propose slimmable compressive autoencoders (SlimCAEs), where rate (R) and distortion (D) are jointly optimized for different capacities. Once trained, encoders and decoders can be executed at different capacities, leading to different rates and complexities. We show that a successful implementation of SlimCAEs requires suitable capacity-specific RD tradeoffs. Our experiments show that SlimCAEs are highly flexible models that provide excellent rate-distortion performance, variable rate, and dynamic adjustment of memory, computational cost and latency, thus addressing the main requirements of practical image compression.
UR  - https://openaccess.thecvf.com/content/CVPR2021/papers/Yang_Slimmable_Compressive_Autoencoders_for_Practical_Neural_Image_Compression_CVPR_2021_paper.pdf
L1  - http://refbase.cvc.uab.es/files/YHC2021.pdf
UR  - http://dx.doi.org/10.1109/CVPR46437.2021.00496
N1  - LAMP; 600.120
ID  - Fei Yang2021
ER  -