PulseAugur
LIVE 10:54:29
research · [3 sources] ·
0
research

Apple researchers unveil learned image codec with 3x bitrate savings and fast mobile speeds

By PulseAugur Editorial · Summary by gemini-2.5-flash-lite from 3 sources

Researchers have developed a new learned image codec that significantly improves the trade-off between perceptual quality and runtime speed. Through a comprehensive study of modeling choices and performance-aware neural architecture search, they created a codec that achieves 2.3-3x bitrate savings compared to traditional standards like AV1 and VVC. This new codec also demonstrates impressive on-device performance, encoding and decoding 12MP images on an iPhone 17 Pro Max in under 300ms. AI

Summary written by gemini-2.5-flash-lite from 3 sources. How we write summaries →

IMPACT This research could lead to more efficient image compression techniques, impacting media storage and streaming.

RANK_REASON This is a research paper detailing a new method for learned image compression.

Read on arXiv cs.CV →

COVERAGE [3]

  1. Apple Machine Learning Research TIER_1 ·

    What Matters in Practical Learned Image Compression

    One of the major differentiators unlocked by learned codecs relative to their hard-coded traditional counterparts is their ability to be optimized directly to appeal to the human visual system. Despite this potential, a perceptual yet practical image codec is yet to be proposed. …

  2. arXiv cs.LG TIER_1 · Kedar Tatwawadi, Parisa Rahimzadeh, Zhanghao Sun, Zhiqi Chen, Ziyun Yang, Sanjay Nair, Divija Hasteer, Oren Rippel ·

    What Matters in Practical Learned Image Compression

    arXiv:2605.05148v1 Announce Type: cross Abstract: One of the major differentiators unlocked by learned codecs relative to their hard-coded traditional counterparts is their ability to be optimized directly to appeal to the human visual system. Despite this potential, a perceptual…

  3. arXiv cs.CV TIER_1 · Oren Rippel ·

    What Matters in Practical Learned Image Compression

    One of the major differentiators unlocked by learned codecs relative to their hard-coded traditional counterparts is their ability to be optimized directly to appeal to the human visual system. Despite this potential, a perceptual yet practical image codec is yet to be proposed. …

RELATED ENTITIES

RELATED TOPICS