Computer Vision News - August 2021

Camila González 25 models and methods that can be used for out-of-distribution detection or uncertainty estimation. As well as having a method that generalizes to out-of-distribution data, you would use all these different methods because you need these extra quality checks .” Previous research exists based on classification models, but Camila says it was a challenge to implement existing methods for semantic segmentation and make sure that they are working properly. There are two proxy tasks in the paper – contrastive learning and edge detection – chosen because they encourage the learning of geometric information that is relevant for the segmentation models. “ We weren’t sure of course when doing the experiments what the results would look like, ” Camila tells us. “ We had a strong assumption because it’s quite a basic expectation that the models will also fail at proxy tasks, but the most exciting part was when we got good results!” There was one challenge that Camila was not expecting, but which could be a lesson for younger researchers out there. “Two of the datasets that we used in our work are from the M&Ms Challenge , and I didn’t realize until quite late in the day that we could ask the authors for additional details of this data,” she reveals. Best of M I D L