I wrote twice earlier from the contact form, didn’t hear back, yet. So posting here in the forum. Please advise on the right place to send if this is not appropriate.
My Background :
My name is Shakti Dhar, I work (and lead) Acumen Software. https://www.linkedin.com/in/shaktidhar
For last 4 years , I have been working in AI and Robotics. Currently I am in process of developing a mobile autonomous robotic assistant.
This will be a robot assistant on wheels (initially. In due course, it will be a humanoid) that uses vision to make sense of the world around it and make decisions.
Although most of my work has been based on self learning, I have done several courses as well, including specialization in Robotics from UPenn.
My work in autonomous algorithms is hitting a severe bottleneck due to lack of available hardware. Boards like Pi are not sufficient, even when setup for parallel processing, Odroid etc are ok but not good enough for real time processing.
Jetson TX looks very promising. A class friend from one of the Robotics courses is already working on it and first level tests were very promising.
But in terms of price, and the availability in India, it’s at the prohibitive side given I am still in experimentation phase.
Purpose of the mail :
I am writing to check what kind of provisions you have in the developer program to assist with hardware. If there is a low cost option or a lease option that I can try before making the final call?
Or anything else you can suggest that can help me evaluate Jestsons TX or similar of your platform before proceeding on it?
I know this has been a long email. Before looking elsewhere, I wanted to give a shot at going with the best I have been able to spot as yet.
Thank you for your time and consideration.
Hi there, Rashi Peripherals is the distributor of Jetson TX1 in India who you may want to contact. If you are an AI startup, see the NVIDIA Inception program. If you are looking for research Academic Seeding, see here for the Hardware Grant Request form: https://developer.nvidia.com/academic_gpu_seeding.
I don’t have an edu email, so not sure if this grant request applies for my case.
Will try though.
Ok, No, it doesn’t seem applicable for my case. It’s specifically for University.
Is there any such program for independent researchers or such?
Although the grant request forms contain verbage inclusive of “Researchers”, and they ask for a university - perhaps you could fill it out with where you graduated from.
What algorithms are you trying to test?- Are you maybe willing to profile a unit remotely over SSH or VNC?
Ok, I will try with the graduation University. Thanks for the tip.
I am facing most limitations in using Convolutional Neural Networks for the purpose of Object recognition and individual identification.
Given the robot is mobile and will be interacting with humans in real time, accuracy, in as small time as possible, is of critical importance.
I need to integrate a camera with the unit to be able to put the system to work. That itself will be one more area where I will have to experiment with several that can work with your platform. The camera for both Pi and Odroid were different, and for some reasons, not talk to each other.
So, working remotely may not be practical. But if all doors close, I might come back and knock this one to try out. Thank you for suggestion.
Ok, I went ahead and applied.
I have been self financing the project so far so never had to write a formal research grant proposal.
Will wait to hear from you.
This looks great. Is this the right forum for the Inception program or there is another forum for it?
There is no specific forum for the NVIDIA Inception program, any issues resulting from the use of NVIDIA hardware or software acquired through the program should go to their respective forums.
Thank you for the response.
I am trying to reach out to someone in Inception Program to check about eligibility and seek a few clarifications.
Any idea how can I reach them?
Will I hear back when my application is processed?
Or I will be informed only if I qualify?
I will appreciate if I can get an idea so that I can plan next steps accordingly.
the email id provided for inception program doesn’t work.
This message was created automatically by mail delivery software.
A message that you sent could not be delivered to one or more of its
recipients. This is a permanent error. The following address(es) failed:
host hqemgate09.nvidia.com [22.214.171.124]
SMTP error from remote mail server after RCPT TO:firstname.lastname@example.org:
550 #5.1.0 Address rejected.
Reporting-MTA: dns; md-in-11.webhostbox.net
Remote-MTA: dns; hqemgate09.nvidia.com
Diagnostic-Code: smtp; 550 #5.1.0 Address rejected.
Apologies; I can verify that the email address posted on the page does not appear to be accessible, we will investigate this further.
Is there any way to know if and when application will be considered or has been processed?
I have hit a wall and can proceed only after I can figure out the hardware part.
I tried a real life application of reinforcement learning in quality assurance of a manufacturing process. Initial attempts of applying learning after 2 to 5 days of processing were not satisfactory (< 65% ). But on gut feeling, I let the training run for 2 weeks, results are at 83%.
Any idea on waiting period will be appreciated. I have dug around, there is no one else in the market to look to.
Thank you for your responses and consideration.
The email address for inception should be working now, please help to give it a try.
I resent the mail. It hasn’t bounced back yet. :-)