TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Video Vectorization

319 点作者 xanthine将近 5 年前

33 条评论

dharma1将近 5 年前
This why Flash was such a nice format for some things (not going into the pitfalls of Flash - but in some ways it was fantastic).<p>I guess these days we have animated SVG, and <a href="https:&#x2F;&#x2F;lottiefiles.com&#x2F;" rel="nofollow">https:&#x2F;&#x2F;lottiefiles.com&#x2F;</a> is getting some traction - but these require you to export in a specific format of course, you can&#x27;t just convert&#x2F;trace a bitmap movie with these. And SVG or Lottie aren&#x27;t designed for longer&#x2F;streaming vector animations, and they don&#x27;t carry synchronised or streaming audio - Flash did all of those things.<p>Vectorisation of bitmap images does have some artifacts, as is evident in the Simpsons demo on this website - when possible, you should export in a vector movie format directly from the vector animation software.<p>It is kind of depressing that we don&#x27;t have an open standard for vector movies (with sound), over a decade after Flash was killed. Sometimes it feels like technology stops or moves backwards.
评论 #23900199 未加载
评论 #23901478 未加载
teddyh将近 5 年前
Isn’t this basically the same thing which was famously used to achieve full motion video in the 1992 Amiga demo “<i>State of the Art</i>”¹ and improved one year later in the followup “<i>9 Fingers</i>”²?<p>1. <a href="https:&#x2F;&#x2F;www.pouet.net&#x2F;prod.php?which=99" rel="nofollow">https:&#x2F;&#x2F;www.pouet.net&#x2F;prod.php?which=99</a> <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=J2r7-ygXOzo" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=J2r7-ygXOzo</a><p>2. <a href="https:&#x2F;&#x2F;www.pouet.net&#x2F;prod.php?which=100" rel="nofollow">https:&#x2F;&#x2F;www.pouet.net&#x2F;prod.php?which=100</a> <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=tGetanBEKK8" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=tGetanBEKK8</a>
评论 #23895894 未加载
评论 #23895753 未加载
评论 #23896040 未加载
评论 #23895781 未加载
emmanueloga_将近 5 年前
It just boggles my mind that the front page of a page that offers a &quot;patented vector-transcoder converts video to a vector format, reducing bitrates&quot; doesn&#x27;t include such video on its front page.
评论 #23895807 未加载
评论 #23896324 未加载
black_puppydog将近 5 年前
This reminds me of a paper from 2005 by Daniel Sýkora et al. [1] which tries something very similar, with the specific use case of animation video. The authors describe it best in the abstract I think:<p>&gt; Video Codec for Classical Cartoon Animations with Hardware Accelerated Playback<p>&gt; We introduce a novel approach to video compression which is suitable for traditional outline-based cartoon animations. In this case the dynamic foreground consists of several homogeneous regions and the background is static textural image. For this drawing style we show how to recover hybrid representation where the background is stored as a single bitmap and the foreground as a sequence of vector images.<p>The idea of using prior knowledge about the nature of the content to decide on an encoding scheme makes intuitive sense to me, though I&#x27;m not a codec person so I don&#x27;t know how feasible it would be to make these ideas into the hardware-accelerated codecs we know from other methods.<p>Of course, these methods would make the <i>most</i> sense when used directly by the animation studios during export, not as an afterthought. But I&#x27;ll take what we can get.<p>By the way, the corpus of Sýkora&#x27;s works [2] is really really impressive in my opinion. He gave a talk in my institute while I was researching methods around neural style, and his take on parametric models, paired with the quality (and speed!) of his results, really left a mark, if not to say they made me seriously question wtf I was doing there. His work is strictly tailored to a professional animation &#x2F; video production setting, so it seems <i>extremely</i> applicable compared to the toy-like nature of neural style methods. That is not to say he doesn&#x27;t know about those. His team&#x27;s recent papers actually fruitfully combine the two.<p>[1]: <a href="https:&#x2F;&#x2F;link.springer.com&#x2F;chapter&#x2F;10.1007&#x2F;11595755_6" rel="nofollow">https:&#x2F;&#x2F;link.springer.com&#x2F;chapter&#x2F;10.1007&#x2F;11595755_6</a><p>[2]: <a href="https:&#x2F;&#x2F;dcgi.fel.cvut.cz&#x2F;home&#x2F;sykorad&#x2F;" rel="nofollow">https:&#x2F;&#x2F;dcgi.fel.cvut.cz&#x2F;home&#x2F;sykorad&#x2F;</a>
评论 #23897107 未加载
gardaani将近 5 年前
Domain specific video formats could be the future. They can reduce bandwidth requirements and offer features not available in pixel videos.<p>One good example is <a href="https:&#x2F;&#x2F;asciinema.org" rel="nofollow">https:&#x2F;&#x2F;asciinema.org</a> , which plays back terminal sessions. The text in the &quot;video&quot; is selectable!
评论 #23900148 未加载
bkm将近 5 年前
One overlooked usecase is RDP and VNC-like concepts. Right now Appetize (App streaming service) uses raster graphics. Vector would allow high-FPS streaming of Apps, making running Apps in the cloud a realistic option.
评论 #23896126 未加载
评论 #23895834 未加载
评论 #23895854 未加载
londons_explore将近 5 年前
Is there market demand for this?<p>In a world where videos are still sent around as multi-megabyte gif files, and audio clips are still distributed with a random slideshow on YouTube, I think a lot of users aren&#x27;t so bothered about efficiency - they just want the simplest thing that works.<p>Lots of music is most efficiently stored as MIDI, yet how many songs on iTunes are midi?<p>For video, raster is king because it works for everything.
评论 #23897148 未加载
评论 #23896338 未加载
评论 #23895811 未加载
评论 #23896023 未加载
评论 #23902506 未加载
评论 #23895547 未加载
评论 #23895525 未加载
评论 #23895794 未加载
评论 #23895616 未加载
royjacobs将近 5 年前
Reminds me of the Spaceballs demos State of the Art and Nine Fingers [1]. Released in 1993 iirc.<p>[1] <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=PPoYzwib7JQ" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=PPoYzwib7JQ</a>
acd将近 5 年前
I think your technology would be useful for restoring old animated videos. Plus it would be useful for The catchy intro animations used on startups to demonstrate their technology.<p>Also would be nice to use when you have bad internet connection speed to watch e-learning material animations.<p>For e-learning you may need a hybrid M3u like Playlist approach with video for the presenter and vector graphics for the screen casts.<p>Manga videos would probably also compress well.<p>Children animated videos.<p>What if you would reduce the color space vectorize of ordinary video to for example 8 colors and smooth out the noise to make large flat surfaces could you compress it with vectors?
评论 #23895867 未加载
laserpistus将近 5 年前
We are using a similar concept in production. Rendered videos for online education within programming. It is a limited usecase of course, but brings the same benefits: 1&#x2F;100 bandwidth; crisp rendered text&#x2F;content instead of rasterized; Easy editing; and the content is searchable &#x2F; indexable.<p>Using html gives us some additional benefits in that we can combine rendered with rasterized content. We also get access to a lot more advanced functionality through the browser context.
评论 #23902424 未加载
vhiremath4将近 5 年前
Woah this is incredible. Even with the increased CPU demands, this is a very large bandwidth savings.<p>&gt;In practice however, DRM, streaming , analytics and ad placement also require javascript logic to function in web runtimes, so in real word settings web-video playback can and does use a non-trivial amount of CPU time.<p>I&#x27;m a little skeptical of this claim. No idea how much CPU is used for DRM, but I can&#x27;t imagine it&#x27;s on the order of multiple percentage points.
评论 #23895600 未加载
评论 #23897416 未加载
评论 #23895982 未加载
gruez将近 5 年前
&gt;Our first vectorized proof of concept for animations is a 17 second clip of the Simpsons located here. Keep in mind, our technology is still at a very early stage, and this is much optimization work left to be done.<p>&gt;<a href="https:&#x2F;&#x2F;files.vectorly.io&#x2F;demo&#x2F;v0-2-simpsons-250kbps&#x2F;index.html" rel="nofollow">https:&#x2F;&#x2F;files.vectorly.io&#x2F;demo&#x2F;v0-2-simpsons-250kbps&#x2F;index.h...</a><p>There isn&#x27;t a raster version to compare to, but that looks noticeably worse than what I&#x27;d expect from a raster version. There&#x27;s a lot of artifacting when there&#x27;s motion, and the linework looks.... off.<p>&gt;<a href="https:&#x2F;&#x2F;files.vectorly.io&#x2F;demo&#x2F;khan-20kbps&#x2F;index.html" rel="nofollow">https:&#x2F;&#x2F;files.vectorly.io&#x2F;demo&#x2F;khan-20kbps&#x2F;index.html</a><p>The khan academy one looks much better, although there&#x27;s still some minor artifacting, eg. when the mouse comes close to the &quot;O&quot; in &quot;O_2&quot; changes a bit.
评论 #23898121 未加载
评论 #23899539 未加载
评论 #23898084 未加载
fefe23将近 5 年前
AFAIK MPEG-4 experimented with encoding 3d objects but it never took off. As usual for MPEG they did not specify how to get the 3d data from a scene but how to encode them, actually how to decode them, so that innovation could happen on the encoding side.<p>The idea is so obvious that I would be astounded if this company gets anywhere. I&#x27;d wager many research teams already attempted this and were never heard from again.<p>Also note that video compression is pretty impressive these days. A typical 2 hour 1080p movie compresses down to a handful of GiB. Compare that to a typical 1080p action game which is easily ten times that big, because storing all the meshes and textures takes a lot of space, it turns out.
评论 #23899895 未加载
cphoover将近 5 年前
Why use computer vision when likely most of these animations come from some software that can&#x2F;could output a vectorized video format?<p>You wouldn&#x27;t have any conversion artifacts that way.
评论 #23897960 未加载
katmannthree将近 5 年前
It sounds cool but given that the demo looks like this [0] on my bog-standard Windows 10 PC with Chrome (and Firefox and Edge) I&#x27;m assuming they&#x27;ve still got some bugs to work out... If it&#x27;s working for anyone here I&#x27;d love to see a screen capture of the proper rendering.<p>[o]: <a href="https:&#x2F;&#x2F;i.imgur.com&#x2F;YO42u2C.png" rel="nofollow">https:&#x2F;&#x2F;i.imgur.com&#x2F;YO42u2C.png</a>
评论 #23895598 未加载
评论 #23897277 未加载
评论 #23896244 未加载
评论 #23898178 未加载
wodenokoto将近 5 年前
As a kid, I used to trace frames from the Simpsons in Macromedia Flash (later Adobe, later discontinued) as a way of creating high-res images, so reading this article really hit home for me!<p>While the output of this algorithm (just like my traces) isn&#x27;t as faithful to the source material as, say, H.264 is, the result looks great and has an amazing style.<p>This might be a great target for mobile-first webtoons.
villgax将近 5 年前
Even their job postings on LinkedIn look like somebody anger-wrote it. For an Image Processing role, they&#x27;ve written no ML&#x2F;DL engineers which is obvious but still, as if Computer Vision isn&#x27;t any way linked to whatever it is they are trying to do with compression.
fredley将近 5 年前
I actually love the slightly &#x27;off&#x27; aesthetic on the simpsons video, and I think there&#x27;s some interesting creative space where this algorithm is deliberately de-tuned for interesting results.
okaleniuk将近 5 年前
It&#x27;s awesome, but it also brings back memories of the flash-animation.
评论 #23896408 未加载
zcw100将近 5 年前
Sounds like <a href="https:&#x2F;&#x2F;github.com&#x2F;fogleman&#x2F;primitive" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;fogleman&#x2F;primitive</a> but for video
sneak将近 5 年前
Patenting algorithms is supposed to be impossible.<p>Companies that ignore this and patent the “system and method” for implementing algorithms are being jerks.
评论 #23896547 未加载
rhn_mk1将近 5 年前
Maybe this finally brings back the crispy scaling quality of Flash videos. They seem to be released in lossy raster formats these days.
tabtab将近 5 年前
If the vectors also relatively smoothly morph in time, then monitor-side interpolation (including motion smoothing) wouldn&#x27;t be needed and directors would have more control over how interpolation is done. They&#x27;ve complained about monitors trying to do too much. It seems pixels are becoming obsolete for video and movies.
karteum将近 5 年前
In terms of stream format, it seems that BIFS &#x2F; MPEG 4 part 11 originaly aimed at the same purpose (probably in a more efficient manner than textual SVG), isn&#x27;t it ? <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;MPEG-4_Part_11" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;MPEG-4_Part_11</a>
chmod775将近 5 年前
While this is pretty cool, this naive approach will fail spectacularly for animation that isn&#x27;t just vector graphics, which is most animation.<p>This might have a future as <i>part</i> of a regular video codec, being used when there&#x27;s mostly vector graphics on screen (or just for those areas that are vector graphics).
bane将近 5 年前
I imagine is a posterization preprocessing step would make this simpler and we could have <i>very</i> low bandwidth &quot;video&quot;. If this could be done in real time, it would dramatically lower the bandwidth required for two-way video chat.
xanthine将近 5 年前
Their android SDK (no release version yet) is available at their Github repo, and so are their bulk upload tools (for talking to their servers using, I guess, a pay to use API).
nautical将近 5 年前
If I understand this correctly, will it also allow to send updates to video on the fly ? Example : Change video from X to Y time to new vectors [V1 ..] ?
villgax将近 5 年前
Not ready yet, just hyping SVG videos for now
vslira将近 5 年前
That&#x27;s really interesting, excited to follow the project and see more!
jcims将近 5 年前
Curious how ML would train on vectorized video instead of rasterized.
imtringued将近 5 年前
Can this technique be used to generate a CAD sketch from a photograph?
评论 #23896606 未加载
iworkfromhome将近 5 年前
This is really cool. So the future video inspiration that will be built using only the code, without shooting again. Because everything can only be made with code. Cool!