Ryan Castellucci :nonbinary_flag: on Nostr: Hacker News commenter: Use vision-capable LLMs to understand if JPEGs and RAWs are ...
Hacker News commenter: Use vision-capable LLMs to understand if JPEGs and RAWs are technically correct but visually corrupt, then automatically recover from tape!
...or hash the original data and use error correcting codes to recover data?
Published at
2024-05-18 15:50:05Event JSON
{
"id": "502cd09e7f0a38846680d9cb0c84d29cbd9ead36419e8975b24ec15fd66cef6e",
"pubkey": "fc74da08b1143a16cdaaf77e54a772e9fb0bfcd1adb79afa3888c9a1e3a742c7",
"created_at": 1716047405,
"kind": 1,
"tags": [
[
"proxy",
"https://infosec.exchange/users/ryanc/statuses/112462882751291973",
"activitypub"
]
],
"content": "Hacker News commenter: Use vision-capable LLMs to understand if JPEGs and RAWs are technically correct but visually corrupt, then automatically recover from tape!\n\n...or hash the original data and use error correcting codes to recover data?",
"sig": "269ce0caf17ddfcfbaa762343b18e024b554e8402904fec14d8ae32e3efae3479efcc7898a24a189c9d8c669a08f92d3b08fcb002881a89dd6ccd110bbaab9b4"
}
