3 ms·
If we're going to go that route, linking to the actual paper might be even more appropriate. Even the blog article is only a second hand source. http://arxiv.o
by slashcom 12y ago
If we're going to go that route, linking to the actual paper might be even more appropriate. Even the blog article is only a second hand source.
http://arxiv.org/pdf/1404.3840v1.pdf http://arxiv.org/pdf/1404.3840v1.pdf
If you read it carefully, there's a caveat about how this particular dataset (recognition of /cropped/ pictures of /unfamiliar/ persons) has relatively low human accuracy (97.5% as opposed to 99.2%), because humans also use other features. That is, there's more work to be done, and facial recognition isn't a "solved" problem yet.
All the same, very impressive work. Congratulations to the authors on achieving such an important milestone.
- dang 12y agoIt's true that the paper is the original source, and it's certainly ok to post those. But a user pointed out recently [1] that many HN users might have time to read a good general-interest article but not the paper itself—and usually the paper is referenced in the thread for those who want it. So either is ok, but if you're going to post a general-interest article, try to make it the most substantive one out there. [1] https://news.ycombinator.com/item?id=7625534 https://news.ycombinator.com/item?id=7625534