TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Extreme video compression with prediction using pre-trainded diffusion models

144 pointsby john_gover 1 year ago

13 comments

Animatsover 1 year ago
Extreme compression will be when you put in a movie and get a SORA prompt back that regenerates something close enough to the movie.
评论 #39436609 未加载
评论 #39438904 未加载
评论 #39435743 未加载
评论 #39438480 未加载
评论 #39437122 未加载
评论 #39441663 未加载
评论 #39436028 未加载
评论 #39438454 未加载
ToJansover 1 year ago
Ahhh, Sloot&#x27;s digital coding system [1] is finally here ;).<p>[1] <a href="https:&#x2F;&#x2F;en.m.wikipedia.org&#x2F;wiki&#x2F;Sloot_Digital_Coding_System" rel="nofollow">https:&#x2F;&#x2F;en.m.wikipedia.org&#x2F;wiki&#x2F;Sloot_Digital_Coding_System</a>
评论 #39437134 未加载
userbinatorover 1 year ago
How fast is this and how big is the decoder&#x2F;encoder? The model weights are not accessible.<p>From the description, it looks like it&#x27;s only being tested with 128x128 frames, which implies that the speed is very low.
评论 #39436454 未加载
IshKebabover 1 year ago
&gt; It can be observed that our model outperforms them at low bitrates<p>It can? Maybe I&#x27;m misunderstanding the graphs but it doesn&#x27;t look like it to me?
评论 #39435808 未加载
holodukeover 1 year ago
Back in 2005 there was a collegue at my first job writing video format converters software. He was considered a genius and the stereo type of an introvert software developer. He claimed that one day an entire movie could be compressesed on a single floppydisk. Everybody laughed and thought he was weird. He might be right after all.
评论 #39441166 未加载
评论 #39438882 未加载
resolutebatover 1 year ago
Here&#x27;s the research behind this: <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;html&#x2F;2402.08934v1" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;html&#x2F;2402.08934v1</a><p>As a casual non-scholar, non-AI person trying to parse this though, it&#x27;s infuriatingly convoluted. I was expecting a table of &quot;given source file X, we got file size Y with quality loss Z&quot;, but while quality (SSIM&#x2F;LPIPS) is compared to standard codecs like H.264, for the life of me I can&#x27;t find any measure of <i>how</i> efficient the compression is here.<p>Applying AI to image compression has been tried before though, with distinctly mediocre results: some may recall the Xerox debacle about 10 years, when it turned out copiers were helpfully &quot;optimizing&quot; images by replacing digits with others in invoices, architectural drawings, etc.<p><a href="https:&#x2F;&#x2F;www.theverge.com&#x2F;2013&#x2F;8&#x2F;6&#x2F;4594482&#x2F;xerox-copiers-randomly-replacing-numbers-in-documents" rel="nofollow">https:&#x2F;&#x2F;www.theverge.com&#x2F;2013&#x2F;8&#x2F;6&#x2F;4594482&#x2F;xerox-copiers-rand...</a>
评论 #39437976 未加载
评论 #39438024 未加载
评论 #39438831 未加载
sbalamuruganover 1 year ago
It’s uncanny how much of the current stuff has been predicted by the sitcom -“Silicon Valley”
评论 #39435895 未加载
评论 #39438706 未加载
LeoPantheraover 1 year ago
It&#x27;s important to remember that any compression gains must include the size of the decompressor which, I assume, will include an enormous diffusion model.
评论 #39438327 未加载
smerikover 1 year ago
Does anyone remember the <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Sloot_Digital_Coding_System" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Sloot_Digital_Coding_System</a>?
zaptremover 1 year ago
Can you share example videos?
评论 #39435104 未加载
hulituover 1 year ago
&gt; Extreme video compression with prediction using pre-trainded diffusion models<p>Is this more extreme than youtube ?
mjevansover 1 year ago
I wonder how effective a speed focused variation could be for quality among 264, 265, and AV1.
hosejaover 1 year ago
Middle-out.