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。

HyperNeRF

288 点作者 montyanderson超过 3 年前

14 条评论

sanxiyn超过 3 年前
Since neural networks are universal, we can often solve problems we don&#x27;t know exactly how to solve with neural networks. NeRF is a great example. But once solved, we should try to reverse engineer the solution and optimize.<p><a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;2112.05131" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;2112.05131</a> did such work and found you can get NeRF quality without any neural network at all and as a result 100x faster.
评论 #29700805 未加载
评论 #29698735 未加载
评论 #29698839 未加载
评论 #29699972 未加载
评论 #29706876 未加载
visarga超过 3 年前
NeRF&#x27;s were inveted in 2020 and started an avalanche of papers (637 citations so far).<p>&gt; Original one: &quot;Nerf, Representing scenes as neural radiance fields for view synthesis&quot;<p>&gt; <a href="https:&#x2F;&#x2F;scholar.google.com&#x2F;scholar?cites=9378169911033868166&amp;as_sdt=2005&amp;sciodt=0,5&amp;hl=en" rel="nofollow">https:&#x2F;&#x2F;scholar.google.com&#x2F;scholar?cites=9378169911033868166...</a><p>What I find most exciting about it is that a NeRF represents images as neural nets, one neural net for each image (in the OP paper generalised to image + deformations). By evaluating the net at various pixel coordinates it gives the color.<p>Up until now learning to replicate the input exactly was called overfitting and considered a bug, not a feature, but they showed a completely new way to wield neural nets.<p>An interesting detail is that they depend on Fourier encoding for the input coordinates. A variant called SIREN uses `sin` as activation function throughout the net.<p>Maybe neural nets will become the data compressors of tomorrow? Shoot a picture, send a neural net around. Game assets could be NeRFs.
评论 #29700106 未加载
评论 #29700736 未加载
评论 #29701463 未加载
themodelplumber超过 3 年前
Wow. I only get the inner workings at a very basic, intuitive level, but it&#x27;s really cool to see the progress of this and similar research. Congrats to the researchers.<p>It&#x27;s awe-inspiring and even frightening at first, in the usual ways, but IMO it has a lot of long-term promise in other ways.<p>Spitballing: I like that this kind of result, which clearly calls into question the role or perception of physical identity, may eventually inform (or even necessitate?) the deconstruction of the physical &quot;I&quot; as a permission broker, and further open a many-to-many interface between the dimensions that underlay what we now think of as &quot;self&quot; and the true depth and variety within what we now think of as &quot;individual humans who are not me&quot;. That opening process alone ought to be a huge jump for human development.<p>Right now we&#x27;re each held, and holding ourselves, way too responsible for maintaining a singular subjective identity, looking at the aggregate. Not only does this compromise our outlook on others based on our subjective perception of the identity match, but it also compromises our ability to reliably consume and metabolize identity-construct-breaking information and experiences. And many of those things, when consumed without so many identity borders--so to speak--will end up being incredibly useful for individuals and group both.<p>Thanks for sharing op.
评论 #29698913 未加载
twoodfin超过 3 年前
In the right circumstances, the NFL would spend $1E7-$1E8 on this or similar tech. It’s wild to think about how much of what we see on screens in a decade or so will be “computationally inferred”.
评论 #29699347 未加载
max_超过 3 年前
Unfortunately, there is a lot of technical Jargon. I don&#x27;t seem to understand much.<p>Could someone help me outline what&#x27;s most interesting here? Maybe applications?
评论 #29701647 未加载
Stevvo超过 3 年前
What a time to be alive!
评论 #29699189 未加载
fxtentacle超过 3 年前
In my opinion, everything nerf related gets a lot of opinion because it&#x27;s highly graphical and thus easy to present. But there&#x27;s few practical applications and it tends to be super slow and not work for more challenging scenes where traditional 20-year old methods like global penalty block matching still work reasonably.<p>And for this paper in particular, I fail to see how they improve over other nerf approaches with deformation terms like Nerfies or D-Nerf
评论 #29700948 未加载
Namidairo超过 3 年前
The team photos double as the demo, that&#x27;s neat. (Mouse over to see the depth colouring.)<p>I presume something along these lines will make it&#x27;s way to the Pixel 6 camera software, given the origin of the research and the onboard edgeTPU block.
bloopernova超过 3 年前
Is this computational &quot;simple&quot; (efficient?) enough that it could be used to create better interactive graphics in things like games?<p>Or am I understanding this incorrectly? Sorry, this feels very much beyond me.
评论 #29701821 未加载
评论 #29703737 未加载
jcims超过 3 年前
What’s the difference between this and using photogrammetry to build a 3D model&#x2F;depthmap and painting it with the images?
评论 #29702185 未加载
评论 #29703769 未加载
fartcannon超过 3 年前
Is there a *Nerf to Blender pipeline?
productceo超过 3 年前
Very interesting. Highly useful for metaverse avatars.
评论 #29699604 未加载
ragmurugesan超过 3 年前
This is cool
ragmurugesan超过 3 年前
Thanks for sharing