Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Interactive deep convnet visualization for Keras and Tensorflow (jakebian.github.io)
124 points by dbranes on Nov 14, 2016 | hide | past | favorite | 11 comments



That looks useful for getting insights into what is happening in a NN.

What is it visualizing in a convolutional layer, does it average all the channels, or select just one?


Each of the channels (filters, in this case) is visualized separately. The user selects a layer and each of the image maps shown represents the activations of that channel.

You can verify this by looking at the `nb_filters` key in the JSON description of the layer on the left and counting the amount of image maps on the right.


As slick as this kind of thing looks, I don't think it does help much to see raw filters on an image-by-image basis. The kind of insights that I think are more important are the abstract ones, like the recent paper about universal adversarial perturbations and the geometry of the decision boundary.

Visualizations can help, but more in the sense of rapidly iterating through bespoke visualizations to generate and test hypotheses. And to do that efficiently you don't need a slick tool that does one kind of visualization, you need a slick 'grammar' to efficiently build new visualizations with, Bret Victor style.


Do you have a link to the paper you are referring to?


https://arxiv.org/pdf/1610.08401v1.pdf

Sorry, I have to ask.. any relation to Dan Fortunato? Maybe that's just a really common name in Boston.


No relation... I think Fortunato is pretty rare. Also: hey Tali! :)


No... He was referring to me! I'm from Boston...


Pretty sure he was referring to me, not you...


It would be nice to also have a way to visualize filters (eg. https://blog.keras.io/how-convolutional-neural-networks-see-...)


I created an account just to say that this looks really fantastic. Thanks for creating this! I'm looking forward to steady improvements in neural network tooling, and it's people like yourself that keep that train moving forward.


License file missing, but setup.py seems to imply MIT.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: