Sei sulla pagina 1di 2

I see that there is only one Topic that a team can select for Hack project,

but the design that we would like to develop involves technologies from 2 topics
(out of 5 topics) listed for this hack.

They are:

1. Machine Learning Application

2. User Interface/HMI

Idea:

Design a web UI which can auto-align based on the device in which the html page is
opened.
(for e.g. mobile phone, PC, tab etc). In the remote end, we will have a server
running in Raspberry PI ARM board.
Devices like robotic arm, Lights etc.. are interfaced to Raspberry pi board so that
they can be controlled via
any above mentioned device which has a route to this ARM board. Also, a camera
connected to ARM would transmit
live feed which can be view in the web UT designed for the above mentioned devices.
Coming to Machine Learning topic, we would use 'speech recognition' + 'text to
speech' technology (libraries),
together as another HMI to interact with ARM board and control robotic arm, switch
on/off electrical appliances etc.
The intention is to build a system like 'IronMan Jarvis' that would interact with
the user and also,
perform actions based on the input given by the user.

In short, we have 2 HMI interfaces to control devices remotely as well as locally +


a live video feed if the
devices were to be controlled remotely.

#######################################################################
# Further Enhancements To Be Done: #
#######################################################################

Speech recognition working commands:

1. record speech.
2. play recorded speech.

https://github.com/xploiter-projects/customer-service-chatbot/blob/master/bot.py

https://www.programcreek.com/python/example/107722/speech_recognition.WavFile

string comparision in speech need not be accurate. even if there is a substring


match, take that as valid input sentence.

Potrebbero piacerti anche