Deep Perceptual Compression
Yash Patel, Srikar Appalaraju, R. Manmatha

TL;DR
This paper introduces Deep Perceptual Compression (DPC), a new deep learning-based image compression method optimized with a perceptual metric aligned with human vision, resulting in visually superior images and better object detection accuracy.
Contribution
The paper proposes DPC, a novel compression technique that jointly optimizes a deep perceptual metric and MS-SSIM, outperforming previous methods and conventional codecs in visual quality and task performance.
Findings
DPC produces visually better images than previous deep learning methods and JPEG-2000.
DPC achieves comparable quality to BPG in human evaluations.
Images compressed with DPC improve object detection accuracy.
Abstract
Several deep learned lossy compression techniques have been proposed in the recent literature. Most of these are optimized by using either MS-SSIM (multi-scale structural similarity) or MSE (mean squared error) as a loss function. Unfortunately, neither of these correlate well with human perception and this is clearly visible from the resulting compressed images. In several cases, the MS-SSIM for deep learned techniques is higher than say a conventional, non-deep learned codec such as JPEG-2000 or BPG. However, the images produced by these deep learned techniques are in many cases clearly worse to human eyes than those produced by JPEG-2000 or BPG. We propose the use of an alternative, deep perceptual metric, which has been shown to align better with human perceptual similarity. We then propose Deep Perceptual Compression (DPC) which makes use of an encoder-decoder based image…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdvanced Image Processing Techniques · Image and Signal Denoising Methods · Image and Video Quality Assessment
