Brian Mount

There's a good source for California accident data (which probably has most of the street typologies of interest) here: http://www.tims.berkeley.edu/ (signup is annoying but automatic, I can send copy or...

I got some great data from Paul, the source files for the sfbetterstreets.org maps are now here: http://sfgeo.org/data/contrib/street-types/

Just saw an interesting instance of this kind of communication that might be food for thought: http://a841-tfpweb.nyc.gov/4thave/what-needs-improving-on-4th-ave-in-park-slope/ This also led me to imagine alternate street sections flowing together in a...

I think in the case of SqueezeNet (where I also got that error), the problem is in the global average pooling layer here: https://gist.github.com/bmount/62089f03e998dc945d9fdb76d8cf82cd#file-squeeze_1-1-prototxt-L652 Does TensorFlow have a generic global...

@arsakes yeah, you can just change the last pooling parameter from global to a fixed kernel with size equal to the width/height of the previous layer (iirc it's 14 or...