PulseAugur
EN
LIVE 15:56:04

Meta's AI codec compresses song 1000x for QR code paper storage

A maker has successfully compressed a two-minute song by approximately 1,000 times using Meta's open-source EnCodec AI neural audio codec. The compressed song, reduced from 2.9MB to about 21KB, was then printed onto eight QR codes on a sheet of paper. Playback of the song requires the same neural network used for compression, highlighting a novel method for data storage and retrieval. AI

IMPACT Demonstrates innovative data storage possibilities by leveraging AI codecs for extreme compression, potentially influencing future archival and data transfer methods.

RANK_REASON The cluster describes a novel application of an existing AI model (EnCodec) for data compression and storage, rather than a new release from a frontier lab.

Read on Tom's Hardware →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

Meta's AI codec compresses song 1000x for QR code paper storage

COVERAGE [2]

  1. Tom's Hardware TIER_1 English(EN) · Luke James ·

    Maker compresses a 2.9MB song by 1000x and prints it on paper as eight QR codes — 21KB song is two minutes long, requires a neural network for playback

    The open-source codec Meta released in 2022 converts a waveform into discrete tokens that a matching decoder turns back into audio.

  2. Mastodon — mastodon.social TIER_1 English(EN) · [email protected] ·

    Maker compresses a 2.9MB song by 1000x and prints it on paper as eight QR codes — 21KB song is two minutes long, requires a neural network for playback The open

    Maker compresses a 2.9MB song by 1000x and prints it on paper as eight QR codes — 21KB song is two minutes long, requires a neural network for playback The open-source codec Meta released in 2022 converts a waveform into discrete tokens that a matching decoder turns back into aud…