Fwd: Google Summer of Code 2018

Chandeepa Dissanayake chandeepadissanayake at gmail.com
Mon Apr 23 15:31:10 EDT 2018

---------- Forwarded message ----------
From: Chandeepa Dissanayake <chandeepadissanayake at gmail.com>
Date: Mon, Apr 23, 2018 at 11:42 PM
Subject: Re: Google Summer of Code 2018
To: "Foster, Jeremiah" <JFoster at luxoft.com>

Hi there!

I would love to extend a huge thank you for selecting my project in GSOC
2018. I'm over the moon right now and just to mention, this is my first
ever GSOC program.

If you'd mind, please explain me more about the background out there and
how you'd like to see the project get going.

I would love to directly speak about my project, this has been an idea
whizzing in my mind since the first moment I learned a bit about smart
vehicles. After submitting the proposal I thought that would not suffice to
explain the technical stuff to you though, ultimately it did.

It would be better if we could divide the project into several blocks, as
this is a project with a collection of ideas. I would love to know your
thoughts on this, but I would simply explain what I'm having in my mind
right now. Still I would not get deeper into technical, on how it's being
done, let's scratch the surface for now. When it comes to different
sections of the project, I would dive deeper and speak about my ideas on
the project.

Currently, what I'm thinking of doing is as follows.
1. Wrapping the whole set of API's currently available around something
called "SmartAPI" - this would be some sort of interface between the
intelligent parts and currently existing facilities.
2. Building a basic model of Voice Command System - this would just convert
voice commands to API calls. It would be better to write this from the
beginning to make sure everything fits into "SmartAPI"
3. Training the models of Voice Command Systems to respond to wide variety
of voice commands:
         I would like to clarify the difference between the existing
systems and the specialty of training this model. Existing systems would
only accept certain commands pronounced in the exact same way. For
           example, if we want to switch on the radio system by voice
commands, we would have to say something like "turn on radio" in the exact
same way always. Instead by training this model, there may be many
   ways of asking to do the same thing, turning back on to the above
example, we may say "switch the radio on" OR "get my radio working" OR
"let's listen to some songs", the model would identify what's to be
   done. Basically, this would be the first step in the intelligent system
I've described in my proposal.

C++ would be a great idea, I would dive deeper into them when we discuss
specifically on each section of the project.

Note: I prefer to use Google libraries such as TensorFlow for training
processes etc. If you are having any other idea, I'm totally open to learn

What I've described would be the first stage of the project I'm thinking
of. Please let me know your ideas on this and what are the basic
requirements(if any) you're hoping to be there the final outcome?

I'd love a skype session or anything like that if you do so.

Thank you!
Hoping to hear from you!


On Sat, Apr 21, 2018 at 6:02 PM, Chandeepa Dissanayake <
chandeepadissanayake at gmail.com> wrote:

> Hello there Foster,
> I hope you can remember me. The student from GSOC, Chandeepa Dissanayake,
> placed a proposal "Extension to Voice Command of IVI System" in GenIVI
> alliance.
> It's near the closure for the process of choosing students.
> I thought it's better to hear from you.
> Any updates out there?
> Thank you!
> On Fri, Apr 6, 2018 at 10:40 PM, Chandeepa Dissanayake <
> chandeepadissanayake at gmail.com> wrote:
>> Thank you for the quick response Foster!
>> I tried to be more intuitive about the gist of my idea in the proposal,
>> so I left the technical part to be discussed later though. With your
>> response, I felt that I've done a good job in writing my proposal, but to
>> be honest, it didn't look quite fine to me though. Anyway, as you think
>> that idea is better, I would elaborate my opinion on how I'm attempting to
>> do what I've mentioned in the proposal if I'm going to be selected.
>> Thank you!
>> Looking forward to hear from you!
>> On Fri, Apr 6, 2018 at 10:32 PM, Foster, Jeremiah <JFoster at luxoft.com>
>> wrote:
>>> Hello Chandeepa!
>>> Thanks for your email. Yes, we've received your proposal -- thank you. I
>>> must say that it looks quite ambitious. I think it is fine that you
>>> submitted a new idea, if you have a passion for a particular technology or
>>> area that can be a powerful motivator, so I believe we are open to GSoC
>>> candidates submitting their ideas. I also think you've got some things in
>>> your proposal that would match well with the tasks that we've defined,
>>> particularly the voice control of an IVI system.
>>> Since we're past the final proposal stage, I think we're in a good spot
>>> to review. I will forward your proposal to colleagues here at GENIVI and
>>> we'll discuss. We try to be open and positive toward student proposals, so
>>> I expect GENIVI to react favorably. The automotive domain is quite complex
>>> and working with embedded Linux requires a problem solving mindset as well
>>> as a willingness to investigate new technologies and tools. While it is a
>>> complex and sometimes difficult domain, we aim to help as best we can so
>>> that both the GSoC project and GENIVI are successful.
>>> Best regards,
>>> Jeremiah
>>> On Wed, 2018-04-04 at 18:47 +0530, Chandeepa Dissanayake wrote:
>>> Hi there,
>>> I’m Chandeepa Dissanayake from Sri Lanka and I’m a student in GSOC 2018.
>>> I’ve submitted a proposal for your organization in Google Summer of Code
>>> 2018. It was not completely one of your ideas but some sort of extension to
>>> one of them.
>>> I’d not elaborate all stuff inside the proposal here, but I’d like to
>>> know about the ideas/types of proposals you’d expect from the community.
>>> Usually the project ideas are based on a set of “pre-defined” tasks to be
>>> completed. I’ve always loved to develop anything from scratch. That
>>> compelled me to submit a new idea.
>>> Hopefully, can you please explain me about the expected types of
>>> proposals you look into? I really look forward to enroll with my desire to
>>> introduce deep learning to smart vehicles.
>>> Thank you!
>>> _______________________________________________
>>> genivi-projects mailing list
>>> <a href="mailto:genivi-projects at lists.genivi.org">genivi-projects at lists.genivi.org</a>
>>> <a href="https://lists.genivi.org/mailman/listinfo/genivi-projects">https://lists.genivi.org/mailman/listinfo/genivi-projects</a>
>>> ------------------------------
>>> This e-mail and any attachment(s) are intended only for the recipient(s)
>>> named above and others who have been specifically authorized to receive
>>> them. They may contain confidential information. If you are not the
>>> intended recipient, please do not read this email or its attachment(s).
>>> Furthermore, you are hereby notified that any dissemination, distribution
>>> or copying of this e-mail and any attachment(s) is strictly prohibited. If
>>> you have received this e-mail in error, please immediately notify the
>>> sender by replying to this e-mail and then delete this e-mail and any
>>> attachment(s) or copies thereof from your system. Thank you.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.genivi.org/pipermail/genivi-projects_lists.genivi.org/attachments/20180424/210c96c8/attachment.html>

More information about the genivi-projects mailing list