I’ve also got some new stuff I’m getting ready to contribute, in the following ML/Deep Learning areas:
Some Basic models using Tensorflow stable 1.13 CIFAR-10 image classifier using a CNN ~86% accuracy – obviously different than Inception-v3/v4 and VGG-16 which we currently have available, but just another option and a good standard benchmark for image labels. Softmax LogReg Netflix Movie Review Sentiment Analyzer using TF. I’ll create a Docker for this like the ImageRecog and VideoRecog dockers. this can and should replace the existing OpenNLP one we have. Will be much easier to use and less code. Audio Speaker Identification using TF and K-means – simple Python/Tensorflow based Speaker Identification and Labeling in an audio file. Face Identification model using CNN – face/no face in an image. Will Dockerize. Should be able to get that contributed in the next 2 months. Cheers, Chris From: Sergey Beryozkin <sberyoz...@gmail.com> Reply-To: "dev@tika.apache.org" <dev@tika.apache.org> Date: Wednesday, July 10, 2019 at 7:05 AM To: "<dev@tika.apache.org>" <dev@tika.apache.org> Subject: [EXTERNAL] Re: Merge flow Thanks, I was just curious, hope to start contributing a bit more... Sergey On Wed, Jul 10, 2019 at 1:39 PM Tim Allison <talli...@apache.org> wrote: Y. Although sometimes I flip the order. :D If it matters or if I’m doing something wrong, let me know! On Wed, Jul 10, 2019 at 4:52 AM Sergey Beryozkin <sberyoz...@gmail.com> wrote: > Hi Tim > > What is the current process for merging the fixes ? The fix goes to the > master first and then it is cherry-picked into the branch_1x ? > > Cheers, Sergey >