nn files are too big to deal with

quantombone / exemplarsvm

Ensemble of Exemplar-SVMs for Object Detection and Beyond

http://www.cs.cmu.edu/~tmalisie/projects/iccv11/index.html

MIT License

430 stars 155 forks source link

nn files are too big to deal with #27

Closed quantombone closed 13 years ago

quantombone commented 13 years ago

Need a way to prune the NN files...they are really intractable

quantombone commented 13 years ago

If we take a single top detection per exemplar per flip (only 2 image views per image), this will results in 2N detections per trainval/test image where N = #exemplars.