ProxIQA: A Proxy Approach to Perceptual Optimization of Learned Image Compression

October 19, 2019 Β· Declared Dead Β· πŸ› IEEE Transactions on Image Processing

πŸ‘» CAUSE OF DEATH: Ghosted
No code link whatsoever

"No code URL or promise found in abstract"

Evidence collected by the PWNC Scanner

Authors Li-Heng Chen, Christos G. Bampis, Zhi Li, Andrey Norkin, Alan C. Bovik arXiv ID 1910.08845 Category eess.IV: Image & Video Processing Cross-listed cs.CV, cs.LG Citations 69 Venue IEEE Transactions on Image Processing Last Checked 5 months ago
Abstract
The use of $\ell_p$ $(p=1,2)$ norms has largely dominated the measurement of loss in neural networks due to their simplicity and analytical properties. However, when used to assess the loss of visual information, these simple norms are not very consistent with human perception. Here, we describe a different "proximal" approach to optimize image analysis networks against quantitative perceptual models. Specifically, we construct a proxy network, broadly termed ProxIQA, which mimics the perceptual model while serving as a loss layer of the network. We experimentally demonstrate how this optimization framework can be applied to train an end-to-end optimized image compression network. By building on top of an existing deep image compression model, we are able to demonstrate a bitrate reduction of as much as $31\%$ over MSE optimization, given a specified perceptual quality (VMAF) level.
Community shame:
Not yet rated
Community Contributions

Found the code? Know the venue? Think something is wrong? Let us know!

πŸ“œ Similar Papers

In the same crypt β€” Image & Video Processing

Died the same way β€” πŸ‘» Ghosted