Best paper award in a workshop of CVPR19

Miriam Bellver, Amaia Salvador, Jordi Torres and Xavier Giro-i-Nieto have been awarded with the best paper award at the CVPR19 DeepVision workshop. The work explored how to invest a limited annotation budget for a computer vision task. Specifically, they proposed a semi-supervised method to cut the annotation budget while outperforming other weakly-supervised methods, showing results in semantic and instance segmentation. The work was developed in collaboration with BSC.

They received a graphics card from NVIDIA as a prize.

See the paper