bmc7505

44 ms·

451.
by bmc7505 9y ago
What troubles me is not the permission to read, but to write. While potentially useful for a host of medical applications, this is a scientifically plausible proposal to create technology that will literally have the ability to control mind
452.
by bmc7505 9y ago
There is a pretty intriguing/troubling observation at the end (cf. https://news.ycombinator.com/item?id=14736254 ): If our brains are doing gradient descent, and we can determine the signals used to indicate the gradie
453.
Deep learning in the brain(drive.google.com)
3 points by bmc7505 9y ago|1 comments
454.
Learning Deep Learning with Keras(p.migdal.pl)
5 points by bmc7505 9y ago|0 comments
455.
Distributed MapReduce with TensorFlow(planspace.org)
3 points by bmc7505 9y ago|0 comments
456.
by bmc7505 9y ago
They reported an accuracy of baseline FBCSP and ConvNets on the BCIC dataset around ~68%-70%, which is reminiscent of early WER for speech recognition (prior to the advent of specialized speech architectures). However, when you consider the
457.
by bmc7505 9y ago
Just skimming the code, this is an impressive piece of work, probably one of the most comprehensive open source ASR implementations using deep learning I've seen yet. It certainly looks correct based on the careful documentation (Engli
458.
by bmc7505 9y ago
Nice! As someone who loves to hike, you should add the National Forest Maps too: https://www.fs.fed.us/visit/maps/
459.
by bmc7505 10y ago
Note that the first example compiles without any further changes, producing valid JVM bytecode.
460.
by bmc7505 10y ago
I was surprised to learn that in Kotlin, it is possible to disambiguate overloaded functions based only on their return type. I had no idea the JVM even supports such semantics. [1] [1]: http://stackoverflow.com/q/42916
461.
by bmc7505 10y ago
Interesting talk about Authorship Detection from 28c3: https://www.youtube.com/watch?v=-b0Ta9h62_E
462.
by bmc7505 10y ago
Hi Andrew, congratulations on your result! A few questions, feel free to answer one or any. How close do you think you are to having fully end-to-end models for speech? Are you optimistic we can get speech synthesis to run on mobile devices
463.
by bmc7505 10y ago
If you like IdeaVim and Vimperator, try using a plugin called AceJump [1]. Like Vimperator, you can map the F key in normal mode to activate AceJump, then start typing any visible character or string on the screen. We are working on multi-m
464.
by bmc7505 10y ago
Google Translate does a surprisingly good job of translating that article. It's pretty funny how the publisher (Scientific American) added their own bias in translating the title and certain phrases, presumably to drive more traffic to
465.
by bmc7505 10y ago
https://medium.com/@ArtificialExperience/creativeai-9d4b2346...
466.
by bmc7505 10y ago
This is why we need plausibily deniable encryption. Does anyone know an Android ROM, or jailbreak app that is visually indistinguishable from the lock screen, which can be unlocked to an innocuous home screen?
467.
by bmc7505 10y ago
Thanks for the feedback! Thought about it, although I'd rather the writing speak for itself. I'm just a guy who enjoys writing. If you enjoy reading, you're welcome to follow me on GitHub or Twitter, sometimes I post interest
468.
by bmc7505 10y ago
Isn't that what social networks are for, validating our own self-importance?
469.
by bmc7505 10y ago
Hey HN, author here! Thanks for reading. If you got through this article, I'm interested in hearing your thoughts! Feel free to PM me (@breandan or bre @ ndan .co). You might also enjoy reading about some of other articles I wrote abou
470.
by bmc7505 10y ago
Author here. Passengers take multiple journerys during Spring Festival. After some fact-checking, 2.9B appears to be the total number of passenger journeys [1] (not just train journeys), which is estimated to be ~332 million in 2016 [2], ac
471.
by bmc7505 10y ago
Not available in Sweden.
472.
Trust in Automation(breandan.net)
79 points by bmc7505 10y ago|14 comments
473.
Trust in Automation(breandan.net)
2 points by bmc7505 10y ago|0 comments
474.
by bmc7505 10y ago
The paper is called, "Obligate plant farming by a specialized ant". (doi:10.1038/nplants.2016.181) http://www.nature.com/articles/nplants2016181
475.
by bmc7505 10y ago
> Did the original WaveNet text to speech demo come with a paper or source code? Paper, yes. [1] Source code, no. [1] https://arxiv.org/pdf/1609.03499.pdf
476.
by bmc7505 10y ago
> Do you know any recent work or paper for speech recognition in language teaching area? What you're describing is called "speech verification". Language education is an application I'm personally very interested in,
477.
by bmc7505 10y ago
Apparently, they tried to use the raw audio waveform with the original setup from the WaveNet paper but couldn't get it to train on their TitanX, so they used MFCCs instead. It's not exactly clear why this is the case. "Secon
478.
by bmc7505 10y ago
The CMU Sphinx project as it stands is basically dead. Even though they recently implemented some sequence-to-sequence deep learning techniques for g2p [1], the core stack is still based on an ancient GMM/HMM pipeline, and current stat
479.
by bmc7505 10y ago
A few weeks ago, a deep learning researcher at one of the world's leading speech groups told me off-the-record that offline, human-parity speech recognition would be "coming soon" to mobile devices. Not sure s/he realize
480.
Democratizing Education in China(breandan.net)
1 points by bmc7505 11y ago|0 comments

More ›