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。

Show HN: Banana ripeness classifier (ML webapp)

32 点作者 gcarvalho大约 6 年前

7 条评论

gcarvalho大约 6 年前
Hey, HN! I am colorblind and I can&#x27;t reliably identify when a banana is ripe enough to eat, so I made a machine learning model that does it for me. The model (92% val accuracy, ~12MB) runs on your device (tf.js) and you can upload directly from the camera on mobile.<p>Feedback is always appreciated!<p>Data, code and trained model are available here: <a href="https:&#x2F;&#x2F;github.com&#x2F;giovannipcarvalho&#x2F;banana-ripeness-classification" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;giovannipcarvalho&#x2F;banana-ripeness-classif...</a>
评论 #19303835 未加载
评论 #19304084 未加载
anotheryou大约 6 年前
I tried to guess an album covers decade: <a href="https:&#x2F;&#x2F;uncover-a-covers-decade.onrender.com&#x2F;" rel="nofollow">https:&#x2F;&#x2F;uncover-a-covers-decade.onrender.com&#x2F;</a><p>You are doing the fast.ai course too, right? [edit: looking at your code, no, you did not :)]<p>I chose my challenge to maximize my learning through something challenging enough it might fail. It&#x27;s only right 60-70% of the time.<p>What I learned:<p>- Binning in to decades might not be Ideal. Something from 79 and 80 is more similar than something from 71 and 79. I guess a better approach would have been smaller bins (3 or more per decade) and than rounding the result to the decade.<p>- Be careful how you gather your learning data. I pulled from the google image search with something like &quot;album OR LP cover 20s&#x2F;30s&#x2F;40s...&quot;; this has ~20% &quot;best of the XXs&quot; collections in them where the cover was designed in a much later decade. I should have searched for &quot;1920 OR 1921 OR 1922...&quot;.
dbs大约 6 年前
I just uploaded an image with some portuguese custard pies and the model went bananas.
simongr3dal大约 6 年前
Nice, pretty cool project. It correctly identified some bananas I had just bought as a little green.<p>Any chance for making an avocado identifier, because those are pretty tough to guess wether they are ripe.
评论 #19305035 未加载
morenoh149大约 6 年前
ha I made something similar for weed, <a href="https:&#x2F;&#x2F;chronicsickness.com" rel="nofollow">https:&#x2F;&#x2F;chronicsickness.com</a>
lovestodonothin大约 6 年前
Cool website! Can you go through how you set up the whole thing?
评论 #19310412 未加载
评论 #19310300 未加载
dplgk大约 6 年前
Next, can you develop Not Banana app? I think you could hit it big with that.