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。

Clay Foundation Model: An open source AI model for Earth

39 点作者 jasondavies超过 1 年前

5 条评论

skybrian超过 1 年前
It's puzzling that there's so much documentation, both about how to run it and the internals, while failing to explain what it does. What are the inputs and outputs? It looks like it does something with satellite photos?
评论 #39022518 未加载
simonw超过 1 年前
<a href="https:&#x2F;&#x2F;clay-foundation.github.io&#x2F;model&#x2F;specification.html" rel="nofollow">https:&#x2F;&#x2F;clay-foundation.github.io&#x2F;model&#x2F;specification.html</a><p>&gt; Clay v0 is a self-supervised modified vision transfer model trained on stacks of Sentinel-2, Sentinel-1 &amp; DEM data. It is trained as a Masked Autoencoder (MAE) to reconstruct the original image from a masked image.<p>&gt; Each data entry is a stack of 10 bands of Sentinel-2, 2 bands of Sentinel-1 &amp; 1 band of DEM data. The model is trained with 3 timesteps of data for each location, with a total of 1203 MGRS tiles globally distributed, each of size 10km x 10km. The data was collected from the Microsoft Planetary Computer.<p>&gt; The model was trained on AWS on 4 NVIDIA A10G GPUs for 25 epochs (~14h per epoch) in December 2023.<p>Also useful: <a href="https:&#x2F;&#x2F;clay-foundation.github.io&#x2F;model&#x2F;model_embeddings.html#format-of-the-embeddings-file" rel="nofollow">https:&#x2F;&#x2F;clay-foundation.github.io&#x2F;model&#x2F;model_embeddings.htm...</a><p>&gt; The embeddings file utilizes the following naming convention:<p>&gt; {MGRS:5}_{MINDATE:8}_{MAXDATE:8}_v{VERSION:3}.gpq<p>&gt; Example: 27WXN_20200101_20231231_v001.gpq<p>MGRS is Military Grid Reference System. I believe 5 characters corresponds to about 60 miles by 60 miles.<p>So presumably the result is an embedding vector representing e.g. a specific spot on earth around a specific 3-year period, such that you can compare that embedding vector with other points to get an idea for what&#x27;s similar and maybe categorize them against known vectors for things like had-a-lot-of-deforestation?
评论 #39034531 未加载
binsquare超过 1 年前
This seems very cool but perhaps additional explanations vs non-ai are useful?<p>I&#x27;m having trouble visualizing benefits of using ai as opposed to traditional data analysis for these purposes: - Track changes in forest cover - Allocate funding for environmental justice
throwaway888abc超过 1 年前
Project website: <a href="https:&#x2F;&#x2F;madewithclay.org&#x2F;" rel="nofollow">https:&#x2F;&#x2F;madewithclay.org&#x2F;</a><p>Pleasant landing page - really enjoyed!
评论 #39021505 未加载
hasbot超过 1 年前
What uses are envisioned?