按年被提到
20212021 · 1 次2026
机器学习与深度学习›训练加速与低精度计算·arXiv:2106.10800 · cs.LG

Lossy Compression for Lossless Prediction

Yann Dubois, Benjamin Bloem-Reddy, Karen Ullrich, Chris J. Maddison · 2021
开始阅读 ↗PDFalphaXiv 讨论
被提及 · 2 人首次 2021-06-23 · 最近 2022-04-10
2不同的人
2活跃月份
2讨论串
31被引(OpenAlex)
arXiv 摘要(候选层不写导读)

Most data is automatically collected and only ever "seen" by algorithms. Yet, data compressors preserve perceptual fidelity rather than just the information needed by algorithms performing downstream tasks. In this paper, we characterize the bit-rate required to ensure high performance on all predictive tasks that are invariant under a set of transformations, such as data augmentations. Based on our theory, we design unsupervised objectives for training neural compressors. Using these objectives, we train a generic image compressor that achieves substantial rate savings (more than $1000\times$ on ImageNet) compared to JPEG on 8 datasets, without decreasing downstream classification performance.

候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。

HN 书评

推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及
评论marauder777 · 2022-04

“While not exactly the idea in your post, this recent paper by Dubois, Bloem-Reddy, Ullrich, and Maddison on "Lossy Compression for Lossless Prediction" (https://arxiv.org/abs/2106.10800) kinda gets at this idea. From the abstract: Most data is automatically collected and only ever "seen" by algorithms. Yet, data compressors preserve perceptual fidelity …”

Silly Image Compression Idea原评论 ↗
被提交为 story · 1 次