• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 53 subscribers
  • Views 303 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CCS/TMS320C5535: Permanent TIesr

Vladimir Botchev
Intellectual 410 points
Part Number: TMS320C5535


Tool/software: Code Composer Studio

Hello,

Is it possible to flash a TIesr build onto the EZDSP5535 board, so that when it boots, and is connected only to a power supply, it starts executing TIesr code? And if so, what are the steps to make this happen?

Best regards,

Vladimir

  • 0
    TI__Guru* 84435 points

    Please see if the boot loader application note helps

    http://www.ti.com/lit/an/sprabl7d/sprabl7d.pdf

    Again, i know you heard us say this multiple times, but do continue to note that our ability to support this device, TIesr and hardware is limited , so I am not able to assure you that we will be able to support you through and through on your journey on this one.

    Looks like you are making progress , even with limited support - so hopefully you are able to get to your end goal.

    For any production level projects, you will need to work with company like Sensory for the final solutions. Please do keep that in mind.

    All the best

    Regards

    Mukul 

  • 0 in reply to Mukul Bhatnagar
    Intellectual 410 points

    Hello,

    Thank you for the support. There appear to be no more issues with the TIesr. It did indeed require fixing wrong makefiles, fixing tools use, but nothing related to TIesr per se, except the heap usage for more than 2 KW, which can be traced in step by step debugging, and fixing the relevant architecture files(not the code). I appreciate your suggestion for Sensory, but I do not believe they are any better than HMM for this type of tasks. May be a notch better. There is a plethora of DNN KWS out there, but even Alexa's own fails miserably in its desire to be universal, e.g. wakes to baleka, so I have no real concerns for your HMM engine, which btw is remarkably engineered. It is Rabiner&Juang and more... Even the huge Baidu Deep Speech 2, which runs on the Mathematica implementation for Raspberry PI, makes errors.

    Best regards,

    Vladimir