In-bin grain drying is a perfect tool to deal with unexpected weather on farms all over the world. The downfall of the method is that it takes a lot of personal experience and can be difficult for farmers to devote the attention needed in order to catch the best conditions for drying. This paper proposes an automatic way to control in-bin grain drying in order to reduce the manual monitoring demanded by farmers. The use of model predictive control is tested on the first layer of the drying bin to assess the practicality and performance. Using simulated results from complex equations as the real world system, an approximated model is used to design a controller which yielded great results in driving the moisture content to the reference.

The automatic generation of image captions has received considerable attention. The problem of evaluating caption generation systems, though, has not been that much explored. We propose a novel evaluation approach based on comparing the underlying visual semantics of the candidate and ground-truth captions. With this goal in mind we have defined a semantic representation for visually descriptive language and have augmented a subset of the Flickr-8K dataset with semantic annotations. Our evaluation metric (BAST) can be used not only to compare systems but also to do error analysis and get a better understanding of the type of mistakes a system does. To compute BAST we need to predict the semantic representation for the automatically generated captions. We use the Flickr-ST dataset to train classifiers that predict STs so that evaluation can be fully automated 1 .