arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.08003eess.IVcs.CVcs.LG
收藏

CompressNet:极低比特率下的生成压缩

CompressNet: Generative Compression at Extremely Low Bitrates

University of Michigan(密歇根大学)

Suraj Kiran Raman, Aditya Ramesh, Vijayakrishna Naganoor, Shubham Dash, Giridharan Kumaravelu, Honglak Lee

摘要:

以极低的比特率(< 0.1 bpp)压缩图像一直是一项具有挑战性的任务,因为由于对压缩数据分配的比特数施加的严格限制,重建质量会显着降低。随着在有限带宽下传输大量图像的需求不断增加,将图像压缩到非常小的尺寸是一项至关重要的任务。然而,现有的方法在极低的比特率下是无效的。为了满足这一需求,我们提出了一种名为CompressNet的新型网络,它通过切换预测网络(SAE-SPN)增强了堆叠式自动编码器。这有助于在这些低比特率(< 0.1 bpp)下重建视觉上令人愉悦的图像。我们在Cityscapes数据集上对我们提出的方法的性能进行了基准测试,在极低的比特率下对不同的指标进行了评估,以表明我们的方法优于其他最先进的方法。特别是,与深度学习SOTA方法相比,在0.07的比特率下,CompressNet实现了22%的感知损失和55%的Frechet Inception Distance(FID)。

英文摘要:

Compressing images at extremely low bitrates (< 0.1 bpp) has always been a challenging task since the quality of reconstruction significantly reduces due to the strong imposed constraint on the number of bits allocated for the compressed data. With the increasing need to transfer large amounts of images with limited bandwidth, compressing images to very low sizes is a crucial task. However, the existing methods are not effective at extremely low bitrates. To address this need, we propose a novel network called CompressNet which augments a Stacked Autoencoder with a Switch Prediction Network (SAE-SPN). This helps in the reconstruction of visually pleasing images at these low bitrates (< 0.1 bpp). We benchmark the performance of our proposed method on the Cityscapes dataset, evaluating over different metrics at extremely low bitrates to show that our method outperforms the other state-of-the-art. In particular, at a bitrate of 0.07, CompressNet achieves 22% lower Perceptual Loss and 55% lower Frechet Inception Distance (FID) compared to the deep learning SOTA methods.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑