ChanServ changed the topic of #mlpack to: Due to ongoing spam on freenode, we've muted unregistered users. See http://www.mlpack.org/ircspam.txt for more information, or also you could join #mlpack-temp and chat there.
cjlcarvalho has joined #mlpack
caiojcarvalho has quit [Ping timeout: 252 seconds]
cjlcarvalho has quit [Quit: Konversation terminated!]
cjlcarvalho has joined #mlpack
caiojcarvalho has joined #mlpack
cjlcarvalho has quit [Ping timeout: 240 seconds]
cjlcarvalho_ has joined #mlpack
ayesdie has joined #mlpack
caiojcarvalho has quit [Ping timeout: 244 seconds]
ayesdie has quit [Client Quit]
ayesdie has joined #mlpack
< ayesdie>
Hello, if anyone is up here, I need some help is adding test features for Neighbor Search and Interpolation alternatives for CF. Just a small help.
< ayesdie>
*need some help in adding
ayesdie has quit [Remote host closed the connection]
ayesdie has joined #mlpack
caiojcarvalho has joined #mlpack
cjlcarvalho_ has quit [Ping timeout: 272 seconds]
ayesdie_ has joined #mlpack
ayesdie_ has quit [Client Quit]
ayesdie has quit [Ping timeout: 252 seconds]
vedantbassi has joined #mlpack
cjlcarvalho_ has joined #mlpack
caiojcarvalho has quit [Ping timeout: 272 seconds]
vivekp has quit [Ping timeout: 244 seconds]
vivekp has joined #mlpack
ayesdie has joined #mlpack
travis-ci has joined #mlpack
< travis-ci>
mlpack/mlpack#5568 (master - a824f23 : Shikhar Jaiswal): The build has errored.
caiojcarvalho has quit [Ping timeout: 250 seconds]
caiojcarvalho has joined #mlpack
cjlcarvalho_ has quit [Ping timeout: 240 seconds]
ayesdie has joined #mlpack
ayesdie has quit [Client Quit]
cjlcarvalho_ has joined #mlpack
caiojcarvalho has quit [Ping timeout: 252 seconds]
cjlcarvalho has joined #mlpack
ayesdie has joined #mlpack
ayesdie has quit [Quit: ayesdie]
caiojcarvalho has joined #mlpack
vedantbassi has quit [Ping timeout: 256 seconds]
cjlcarvalho has quit [Ping timeout: 252 seconds]
cjlcarvalho has joined #mlpack
caiojcarvalho has quit [Ping timeout: 244 seconds]
cjlcarvalho_ has quit [Ping timeout: 245 seconds]
cjlcarvalho_ has joined #mlpack
cjlcarvalho has quit [Ping timeout: 252 seconds]
petris_ has quit [Quit: Bye bye.]
petris has joined #mlpack
< davida>
I have a question regarding the correct data layout for input to a Convolution layer. This layer expects inSize, inputWidth and inputHeight. I have a set to 1,080
< davida>
64x64x3 images as the input to the first conv layer.
travis-ci has joined #mlpack
< travis-ci>
mlpack/mlpack#5569 (master - 971d78b : Shikhar Jaiswal): The build has errored.
< davida>
It seems reasonable that inSize=3, inputWidth=64 and inputHeight=64. However, I have to map this data into a matrix of 'm' examples so I need to flatten the 64x64x3 image
< davida>
into a matrix of 12288x1080 (m=1080).
< davida>
The question I have is what layout does the Conv layer expect the 64x64x3 data to be in? Should I write 64x64 row by row then another 64x64 for the second channel?
< davida>
Or does the Conv layer expect the input maps to be laid out in some other fashion?
< davida>
Unfortunately the example provided in the mlpack/models GitHub link only covers a CNN with a single map on the first input so I cannot work out the correct input format for the matrix.
< davida>
Any advice would be really welcomed. Thanks.
< zoq>
davida: It should be flatten (64 * 64) (64 * 64) (64 * 64), you can use arma::vectorise to do the actual reshaping.