| # The blog post that started it all: https://neocities.org/blog/the-fcc-is-now-rate-limited | |
| # | |
| # Current known FCC address ranges: | |
| # https://news.ycombinator.com/item?id=7716915 | |
| # | |
| # Confirm/locate FCC IP ranges with this: http://whois.arin.net/rest/net/NET-165-135-0-0-1/pft | |
| # | |
| # In your nginx.conf: | |
| location / { |
This is an Keras implementation of ResNet-152 with ImageNet pre-trained weights. I converted the weights from Caffe provided by the authors of the paper. The implementation supports both Theano and TensorFlow backends. Just in case you are curious about how the conversion is done, you can visit my blog post for more details.
ResNet Paper:
Deep Residual Learning for Image Recognition.
Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun
arXiv:1512.03385
A work in progress collection of proprietary and as of yet undocumented HomeKit characteristics and their UUIDs used by Elgato Eve.
Based on the work by gomfunkel and 0ff. Characteristics and data dump for Door, Motion and Thermo thanks to @NebzHB.
More infos not yet incorporated in the comment section.
This list is not including all Eve accessories available and some services and characteristics still make no sense to me. If you have anything to contribute, please leave a comment. There is no guarantee that the information listed below is correct.