Nick, Here are some pointers: 1. Image recognition using Tensorflow: https://wiki.apache.org/tika/TikaAndVision; Link to Paper: https://memex.jpl.nasa.gov/MFSEC17.pdf 2. Image Recognition using Deeplearning4j - https://wiki.apache.org/tika/TikaAndVisionDL4J 3. Sentiment Analysis using OpenNLP: https://github.com/apache/tika/pull/169 4. Video labeling using tensorflow image rec: https://wiki.apache.org/tika/TikaAndVisionVideo 5. Named Entity Extraction using OpenNLP and CoreNLP: https://wiki.apache.org/tika/TikaAndNER
*Coming soon (Work in progress):* 6. Image Captioning (Image-to-Text) https://github.com/apache/tika/pull/180 Cheers, -Thamme *--* *Thamme Gowda* TG | @thammegowda <https://twitter.com/thammegowda> ~Sent via somebody's Webmail server! On Tue, May 16, 2017 at 6:59 AM, Chris Mattmann <[email protected]> wrote: > Yep, literally take a look at the Tika wiki – there are examples a plenty > and even > screen shots. Further, if you look at the MEMEX site under our new > publications > section, there are a few examples (like the ICMR paper on forensics) that > show it > in action. > > http://memex.jpl.nasa.gov/#publications > > > > On 5/16/17, 6:21 AM, "Konstantin Gribov" <[email protected]> wrote: > > IIRC, image and video labeling basic support was added (Chris & Thamme > could you elaborate on that, please), TSD (TIKA-2309, time stamped data > envelope format) support, slf4j migration (ongoing on 2.x branch). > > вт, 16 мая 2017 г. в 16:06, Allison, Timothy B. <[email protected]>: > > > Doh! Sorry for the delay...might add configuration of > EncodingDetectors, > > but that's probably too far into the weeds? > > > > -----Original Message----- > > From: Nick Burch [mailto:[email protected]] > > Sent: Sunday, May 14, 2017 11:34 AM > > To: [email protected] > > Subject: Tika talk next week - help needed! > > > > Hi All > > > > Last year in Seville, I gave a talk on Tika entitled "Apache Tika - > What’s > > new with 2.0?". For ApacheCon Miami next week, I've been roped into > giving > > an updated version... > > > > https://apachecon2017.sched.com/event/9zvD/apache-tika- > whats-new-with-20-nick-burch-apache-software-foundation > > > > My slides from Seville are available at: > > > > http://events.linuxfoundation.org/sites/events/files/slides/ > WhatsNewWithApacheTika_1.pdf > > > > Beyond updating the list of releases and parsers, and the slide > > background, what should I change? > > > > Maybe some more on Tika eval? More details on some of the NLP / > Entity > > Recognition / Image Recoginition stuff? Some screenshots of that > stuff? > > More on translation? Something else? > > > > Ideas greatly appreciated! Good screenshots even more so :) > > > > Cheers > > Nick > > > -- > > Best regards, > Konstantin Gribov > > > >
