ChanServ changed the topic of #mlpack to: "mlpack: a fast, flexible machine learning library :: We don't always respond instantly, but we will respond; please be patient :: Logs at http://www.mlpack.org/irc/
ImQ009 has joined #mlpack
< KimSangYeon-DGU[>
kartikdutt18: sakshamb189 Are you there?
< kartikdutt18[m]>
Sure, I'm working on fixing and writing tests for YOLO model so that after that we can write the detection state based on preprocessing.
< kartikdutt18[m]>
I think I should be able to fix a commit for it by tomorrow.
< kartikdutt18[m]>
* I think I should be able to push a commit for it by tomorrow.
< kartikdutt18[m]>
I think the model is complete but we need to get the input format so that we add the detection loss function.
< KimSangYeon-DGU[>
I see, Ok
< kartikdutt18[m]>
So the basic model (like portion with ann layers) is complete. The only thing we need is the loss function for detection stage and for that we need to know the input so that we know what to extract.
< KimSangYeon-DGU[>
Is the input format different from DarkNet?
< kartikdutt18[m]>
For bounding boxes, we will feed feild type as input and generate an output feature map which will encode the input bounding boxes.
< kartikdutt18[m]>
For YOLOv1 the implementation is correct. I have to correctly extend it for v3.
< KimSangYeon-DGU[>
Ok, thanks for the clarification
< KimSangYeon-DGU[>
I'm checking our plan
< KimSangYeon-DGU[>
I think the project goes smoothly. after finishing YOLOv1 and v3, we'll move on the visualization part and that is end of the project.
< kartikdutt18[m]>
Right, I can probably do the visualization part in a day or two. So mainly we have to work YOLOv1 and v3.
< KimSangYeon-DGU[>
Also, I'd like to note that YOLO model will be another nice feature from users, so please let's think about API carefully.
< kartikdutt18[m]>
<KimSangYeon-DGU[ "Also, I'd like to note that YOLO"> That was resolved in the DarkNet PR. For some reason I had to mention to include set explicitly.
< kartikdutt18[m]>
I can merge in the master branch that should fix it.
< KimSangYeon-DGU[>
Ok
< KimSangYeon-DGU[>
Is there anything to discuss more?
< kartikdutt18[m]>
I think that's it from my side. Hopefully if we should be able to merge the YOLO model this week as well.
< KimSangYeon-DGU[>
Yes, I think so
< KimSangYeon-DGU[>
Thanks for the meeting and have a nice week!
< kartikdutt18[m]>
Great, Have a great week.
theloudspeaker has quit [Quit: authenticating]
< rcurtin>
zoq: could you install libcereal-dev on polar and pioneer?
theloudspeaker has joined #mlpack
< zoq>
rcurtin: sure, one second
< rcurtin>
thanks :)
< zoq>
done
< KimSangYeon-DGU[>
I think I should be able to push a commit for it by tomorrow.