Yukkuri on Nostr: Christmassy Merry Dude There is a bunch of whitepapers on dealing with ...
Christmassy Merry Dude (npub1ecj…fchy) There is a bunch of whitepapers on dealing with salt-and-pepper noise up to 0.9 using just general algorithmic approaches, and you can probably go past that with specifically trained network, so until someone actually bothers with implementing at least the 3d reconstruction, any noise might be just excessive, and past that moment - won't stop anyone dedicated to image/shape recognition. I'm thinking on adding some form of Raven's progressive matrices as another module, at least it would be entirely different domain for someone to tackle.
Published at
2024-03-13 07:16:40Event JSON
{
"id": "fb0e23be60ca79a655fb54a096c02d42246263f64d80f78fc6dce54a8e7aa194",
"pubkey": "93e8db96f9dd979cb19accd8a8419a10ff7916057d442b4e1b27c1b505f17541",
"created_at": 1710314200,
"kind": 1,
"tags": [
[
"p",
"ce251da465f8986f3afa4eef42b05ace5715576d5b7e1280c2031dc6051f74cc",
"wss://relay.mostr.pub"
],
[
"e",
"72fa27da2c3d23008e111f4f1abf635c7ddefa50f595923720cb4dfe37a3db16",
"wss://relay.mostr.pub",
"reply"
],
[
"proxy",
"https://eientei.org/objects/9b1c0469-d46d-4c4e-be0d-891f3526f5f6",
"activitypub"
]
],
"content": "nostr:npub1ecj3mfr9lzvx7wh6fmh59vz6eet324mdtdlp9qxzqvwuvpglwnxqv6fchy There is a bunch of whitepapers on dealing with salt-and-pepper noise up to 0.9 using just general algorithmic approaches, and you can probably go past that with specifically trained network, so until someone actually bothers with implementing at least the 3d reconstruction, any noise might be just excessive, and past that moment - won't stop anyone dedicated to image/shape recognition. I'm thinking on adding some form of Raven's progressive matrices as another module, at least it would be entirely different domain for someone to tackle.",
"sig": "3896944feca92e900ef836ba695d109c2a2ef16558546b86a5872b6dc0dac2006c44a71ae6c0c1f611bb1114e9406e3b32c935ca2c2d554969743147b4c11949"
}
