Semantic Segmentation with Second-Order Pooling - Code
Semantic Segmentation with Second-Order Pooling
Joao Carreira, Rui Caseiro, Jorge Batista and Cristian Sminchisescu
Results on a random sample of images from the Pascal VOC 2011 Segmentation test set (48% VOC score, comp6).
J. Carreira, R. Caseiro, J. Batista and C. Sminchisescu. Semantic Segmentation with Second-Order Pooling. European Conference on Computer Vision (ECCV), October 2012. pdf (talk).
The code uses the CPMC segmentation algorithm. If you use it, please cite also:
J. Carreira and C. Sminchisescu. CPMC: Automatic Object Segmentation Using Constrained Parametric Min-Cuts. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012. pdf.
You can now download source code to reproduce the results in the paper: code.
This should work in a Linux 64 bits computer using MATLAB R2011b. For the VOC experiment it is recommendable to have 16 or preferrably 32 GB of RAM
and at least 200GB of disk space.