Automotive Working Group Teleconference

<scribe> scribenick: ted

Ted reviews agenda asks if there are other topics

Mike: I would be interested in voice interaction within automotive context as that is my current focus

Ted: voice has been of considerable interest but unfortunately we were never able to engage Nuance who was a W3C Member. There is a W3C Community Group based on Google's API which some are following from automotive. Can you tell us more about what you are doing at IBM with voice?

Mike: what we are working on is voice interactions with humans, could be from a phone, laptop etc and used as a chat interface
... using natural language understandings
... I have used Nuance's engine before
... we have intent entities and dialogs
... if a user sends a text compute it comes back as voice and in reverse voice to textual intent
... entities could be an object like a day of the week, location, etc that gets extracted and iterpretted
... this is based on the Watson conversation service and there are similar from FB and Google

Hira: we need to clarify use cases for automotive and other uses
... this is partially covered by MMI WG

Ted: I thought MMI more focused on playback, combining subtitles and closed captioning with voice than it is about speech interpretation?

Paul: OEM like the idea of voice command based control

https://www.w3.org/2002/mmi/

Ted: it does include input/speech
... Mike, do you know how much your group had followed MMI?

Mike: not sure, we are looking to apply this within IoT space
... starting to adopt this Watson voice platform
... also visual recognition
... not sure if those groups in IBM are part of W3C

Ted: visual recognition is of great interest as well for recognizing [parts of] signs, pedestrians
... MMI is a closed WG

Ted:visual recognition is also of great interest in the industry but there has been no standards work in W3C Automotive in that space. There will be demos from Renasas and others next week on their recognition engines

Ted: there is a stronger more immediate (if not overdue) need for voice imho. visual recognition is of more interest to the upcoming driver assist and autonomous focus the industry is moving toward

Paul: we had Dominec from @@ talking with various people from Google's CG effort

Ted: we should probably survey the group and see if there is enough interest in voice to try to explore it more in the BG

Paul: as you said it requires critical mass

F2F

Paul: basically I am most interested in covering F2F agenda on this call

https://www.w3.org/2002/09/wbs/61259/uk2017f2f/results

Ted: VISS and VIAS are both close to their respective milestone publications, both have outstanding issues and think since most of the editors are not on this call defer those topics to the F2F and think on other agenda.

Paul: Genivi has settled on browser baked into their platform. Rudi is leading their Networking Expert Group
... we should look at liaisons within Genivi groups to join our F2F

Ted: also Genivi Security Expert Group
... I'll reach out to Stacey

Topics:

-VISS

-VIAS

-Testing

-RSI

-Payments

-Security and Networking group liaisons

-WG recharter

Junichi: regrets for the F2F
... our focus needs to be on VIAS and VISS specifications
... Urata has some topics that touch on security within the specs

<hira> https://w3c.github.io/automotive/vehicle_information_api/vehicle_information_api_specification.html

Paul: to me it boils down to finalizing the tests and spec
... finishing up outstanding issues and open to hearing your plan on how to proceed

Ted: I have seen some things around tokens from Urata
... as for your question on the scope of Genivi's Security Expert Group, they are going to be turning their attention on our work soon

-Future BG incubator TFs

(eg speech)

Ted: I will update the wiki. Junichi and Mike, send topics and availibility for remote participation

Hira: I will join remotely too

Ted: send to member-automotive@w3.org your availablility and interest as well

[adjourned]

[[Mike Aro20:17

The World of Watson enters your car

IBM's Bret Greenstein explains how automobiles will find their voice.http://www.autoblog.com/2017/04/19/world-of-watson-enters-your-car/

]]