built-in functions to create convolutional and maxPooling layers

[Muhammad Aitsam]

Hi,

I want to know that are their built-in functions to create convolutional layer and Max-pooling layer in SpyNNaker?

If yes, then kindly share the relevant document.

If no, what is the best approach to create multi-layer network in SpiNNaker.

Thanks in advance.

Regards,

Aitsam

[Andrew Gait]

Hi,

It depends what sort of convolutional/pooling layer you want to use.  There is support for 2D convolutions as detailed here:

http://spinnakermanchester.github.io/development/2d_convolutions.html

but if you just need 1D convolutions or max pooling you may be better considering the KernelConnector, examples of which can be seen here:

https://github.com/SpiNNakerManchester/sPyNNaker/blob/master/spynnaker_integration_tests/test_connectors/test_kernel_connector.py

Andy

----------------------------------------------------
Dr. Andrew Gait (he/him),

Research Software Engineer,

APT Group,
Department of Computer Science,
The University of Manchester,
Oxford Road, Manchester M13 9PL

email: andre...@manchester.ac.uk
web: http://personalpages.manchester.ac.uk/staff/Andrew.Gait/

I do not expect a reply to this email outside of normal working hours

---

From: spinnak...@googlegroups.com <spinnak...@googlegroups.com> on behalf of Muhammad Aitsam <aitsam...@gmail.com>
Sent: 24 October 2022 12:28
To: SpiNNaker Users Group <spinnak...@googlegroups.com>
Subject: [SpiNNaker Mailing List] built-in functions to create convolutional and maxPooling layers

 

--
You received this message because you are subscribed to the Google Groups "SpiNNaker Users Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to spinnakeruser...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/spinnakerusers/45ab6338-8082-495c-84a7-93f43e63c82bn%40googlegroups.com.