llebttam
Re:Computer Engineering for Babies (2021)
on Thursday January 18, 2024 @03:44AM (#39039459)
So I'm not a morning person. As a new parent of a 1-year-old, I realized I can get an extra 45 minutes of sleep in the morning by leaving books in our baby's crib for her to look through when she wakes up. One night I left Computer Engineering for Babies in there, and after waking the next morning I looked on the baby monitor and I saw her turning the pages of the book, methodically pressing both buttons for the AND gate, one button at a time for the OR gate, etc. It was probably one of my proudest father moments so far. She was about 18 months at that point, and I hadn't seen her really understand the gates before.
Re:Scientists say they can read nearly the whole genome of an IVF-crea...
on Monday March 21, 2022 @06:49PM (#30760764)
Commercial products in the space have been around for several years but can only screen for a small number of genetic diseases at a time. For example, Natera's Spectrum screens for chromosomal abnormalities (eg Down Syndrome) plus one custom genetic problem for a single condition of greatest concern. https://www.natera.com/womens-health/spectrum-preimplantatio...
Some polygenic screening services that look at a much larger number of conditions are just reaching market now (eg Orchid Health), though our ability to map genes correlated with polygenic traits to phenotype results is still in its early stages. https://www.orchidhealth.com/
Re:Virtual tour of the tomb of Pharaoh Ramses VI
on Friday May 01, 2020 @07:32PM (#23047589)
Glad you like it!
Matterport's camera uses a mixture of structured light, deep learning, and other techniques to build a 3D point cloud from each capture. Then, the app aligns the different captures together to produce a 3D mesh.
It's also possible to use consumer 360 cameras with Matterport's app, though accuracy is lower. In this case, the 3D-from-2D estimation primarily happens using deep learning.
You can learn more here: https://matterport.com/how-it-works https://matterport.com/cameras
Matterport is a YC company from way back in 2012.
Re:A global reforestation project is how we fix climate change
on Thursday January 16, 2020 @06:07PM (#22070057)
FWIW, I don't think he's creating a monocrop forest -- he's planting a range of species that form an interdependent web.
This is also being done in locations that were forests a few hundred years ago before humans cut them down. The mature forest is a self-sustaining ecosystem, but it needs extra water for a few decades until it reaches that point. https://www.westhawaiitoday.com/2020/01/12/hawaii-news/yisha...
Re:A global reforestation project is how we fix climate change
on Thursday January 16, 2020 @06:00PM (#22070007)
The world has a lot of land, though. Before humans came along, there were roughly 6 trillion trees in the world. Now there are about 3 trillion, mainly due to forests being cut down for agriculture, wood, etc. Getting back the other 3 trillion trees would sequester an additional 75 gigatons of CO2 per year using your figure of 25kg/tree/year. Human activity emits 45gt/year according to his article.
Re:A global reforestation project is how we fix climate change
on Thursday January 16, 2020 @05:48PM (#22069889)
A friend of mine who's studied climate change in detail recommended the following company for voluntary carbon offsets:
Re:A global reforestation project is how we fix climate change
on Thursday January 16, 2020 @05:06PM (#22069535)
Yeah, it's definitely a lot.
I think the actual costs will be a lot lower as this will take many years of experimentation before scaling, and by then the cost of the solar panels (which I think is the vast majority of the cost) will likely have continued to ride its exponential downward cost curve. One source: https://earthtechling.com/solar-energy-costs-trends/
Re:A global reforestation project is how we fix climate change
on Thursday January 16, 2020 @04:47PM (#22069335)
I agree, but I still think of this as a useful part of a complete solution. It's something that buys us time to shift away from fossil fuels.
With current tech, we could drastically reduce emissions by moving to a mix of nuclear (the modern, safe kind) for baseload plus wind and solar for peak, but it's going to take few decades to build it all out, and during the intervening period we have a lot of excess CO2 we need to sequester.
Re:A global reforestation project is how we fix climate change
on Thursday January 16, 2020 @04:43PM (#22069293)
I'm not an expert in this area, but from what I can tell:
Young rapidly growing forests sequester carbon more quickly than mature forests, but mature forests still sequester some carbon, possibly into the soil. (One source https://www.sciencedaily.com/releases/2008/09/080910133934.h... ) Separately, the regrown forest could be maintained in a state of peak carbon sequestration by doing selective logging from time to time, and as a bonus it would be profitable to do so.
Re:A global reforestation project is how we fix climate change
on Thursday January 16, 2020 @04:29PM (#22069164)
Some thoughts on finite CO2-absorbing capacity:
- If these forests are sustainably logged, we can take some of the carbon out and sequester it in buildings (or even just bury it), allowing the forest to regrow and absorb more carbon.
- I have a theory that presently discussed global climate change solution only needs to buy us a maximum of 100 years, and by then our tech for dealing with silly problems like excess carbon will have almost surely advanced beyond recognition. Given rates of progress in chemistry, physics, and materials science over the last two centuries, this isn't crazy.
Re:Announcing the Matterport3D Dataset of Labeled 3D Spaces for Machin...
on Thursday September 21, 2017 @04:10PM (#15306913)
If you didn't see the link to the paper, it's here: https://arxiv.org/pdf/1709.06158.pdf A wide range of usecases are discussed there.
Re:Announcing the Matterport3D Dataset of Labeled 3D Spaces for Machin...
on Thursday September 21, 2017 @04:09PM (#15306907)
Lots of things! Since 3D convolutional networks are very limited in their maximum resolution, most of the interesting things you can do involve learning on RGB+D images via a 2D CNN. A lot of tasks on images (segmentation, identification etc) are easier when you have even partial depth data as input.
Announcing the Matterport3D Dataset of Labeled 3D Spaces for Machine Learning
Matterport (YC W12) Is Looking for More Senior Back End Engineers
Matterport makes a 3D camera and web platform that allow users to easily capture and display 3D models of physical spaces. Check out some example spaces in our gallery: https://matterport.com/galleryThe Backend Engineering Team works on our Platform and APIs. The stack is Django/Python/Postgres on AWS, and Fastly as a CDN. We're a small team right now, at 4 engineers who mostly do Django and 3...
Explore a WebGL 3D model of YC made with the Matterport (YC W12) 3D camera
Matterport (YC W12) enables you to make online 3D copies of physical places
We're building a 3D reconstruction system that makes it easy for anyone to automatically build 3D models of physical spaces such as houses in just a few minutes. We have an initial version of our product in beta with large companies in home improvement, real estate, construction, and insurance, and things are going very well. You can check out what our reconstructions look like at...
Love computer vision? Come reconstruct the world in 3D at Matterport (YC W12)
Matterport enables you to make online 3D copies of physical places.We're building a 3D reconstruction system that makes it easy for anyone to automatically build 3D models of physical spaces such as houses in just a few minutes. We have an initial version of our product in beta with large companies in real estate, construction, and home improvement, and things are going very well. You can check...
Matterport (YC W12) automatically makes online 3D copies of physical places
We're building a 3D reconstruction system that makes it easy for anyone to automatically build 3D models of physical spaces such as houses in just a few minutes. We have an initial version of our product in beta with large companies in real estate, construction, and home improvement, and things are going very well. You can check out what our reconstructions look like at...
Scan the world in 3D - MatterPort is hiring its first full-stack web developer
Matterport (YC W12) creates a low-cost easy-to-use 3D scanner that lets you turn reality into 3D models.We scanned Y Combinator shortly before demo day -- you can see the scanning process and the automatically generated 3D model on our demo video starting at 0:40 http://matterport.com/#videoWe're proud to be creating something incredibly useful. This is going to become a vital tool for the...
MatterPort - Software Engineer (Vision & Algorithms focus) - 3D Scanning Company
What we're doing:We are creating a low cost point-and-shoot 3D scanner. Our software, combined with the inexpensive 3D capture hardware now available on the market, will allow everyone to create 3D models of physical objects and environments quickly and easily. See a demo video at http://matterport.comWhat you would be doing:You’ll be working on our 3D alignment and mesh reconstruction...
Re:Did Y Combinator send out the acceptance/ rejection emails yet?
on Monday October 31, 2011 @10:14PM (#3180249)
just got one (or rather, we got a "come in for an interview" email)
Re:Realtime Dynamic 3D surface reconstruction and interaction...using ...
on Thursday August 11, 2011 @10:36PM (#2875748)
I've done a lot of work in this area, and I can say that this is significantly faster and higher quality than the other Kinect-based 3D reconstruction techniques out there such as RGBDemo ( http://www.youtube.com/watch?v=Cldf7UdFq1k ).
It's also clear just how much of an advantage having a 3D sensor is for reconstruction when you compare this against 2D-camera-based 3D reconstruction software like Photofly.