Guided super-resolution is a unifying framework for several computer vision tasks where the inputs are a lowresolution source image of some target quantity (e.g., perspective depth acquired with a time-of-flight camera) and a high-resolution guide image from a different domain (e.g., a grey-scale image from a conventional camera); and the target output is a high-resolution version of the source (in our example, a high-res depth map). The standard way of looking at this problem is to formulate it as a super-resolution task, i.e., the source image is upsampled to the target resolution, while transferring the missing high-frequency details from the guide. Here, we propose to turn that interpretation on its head and instead see it as a pixel-to-pixel mapping of the guide image to the domain of the source image. The pixel-wise mapping is parametrised as a multi-layer perceptron, whose weights are learned by minimising the discrepancies between the source image and the downsampled target image. Importantly, our formulation makes it possible to regularise only the mapping function, while avoiding regularisation of the outputs; thus producing crisp, naturallooking images. The proposed method is unsupervised, using only the specific source and guide images to fit the mapping. We evaluate our method on two different tasks, superresolution of depth maps and of tree height maps. In both cases we clearly outperform recent baselines in quantitative comparisons, while delivering visually much sharper outputs.

No abstract

No abstract

Herbarium sheets present a unique view of the world's botanical history, evolution, and diversity. This makes them an all-important data source for botanical research. With the increased digitisation of herbaria worldwide and the advances in the fine-grained classification domain that can facilitate automatic identification of herbarium specimens, there are a lot of opportunities for supporting research in this field. However, existing datasets are either too small, or not diverse enough, in terms of represented taxa, geographic distribution or host institutions. Furthermore, aggregating multiple datasets is difficult as taxa exist under a multitude of different names and the taxonomy requires alignment to a common reference. We present the Herbarium Half-Earth dataset, the largest and most diverse dataset of herbarium specimens to date for automatic taxon recognition.

In the age of big data, scientific progress is fundamentally limited by our capacity to extract critical information. We show that recasting multispecies distribution modeling as a ranking problem allows analyzing ubiquitous citizen-science observations with unprecedented efficiency. Based on 6.7M observations, we jointly modeled the distributions of 2477 plant species and species aggregates across Switzerland, using deep neural networks (DNNs). Compared to commonly-used approaches, multispecies DNNs predicted species distributions and especially community composition more accurately. Moreover, their setup allowed investigating understudied aspects of ecology: including seasonal variations of observation probability explicitly allowed approximating flowering phenology, especially for small, herbaceous species; reweighting predictions to mirror cover-abundance allowed mapping potentially canopy-dominant tree species nationwide; and projecting DNNs into the future allowed assessing how distributions, phenology, and dominance may change. Given their skill and their versatility, multispecies DNNs can refine our understanding of the distribution of plants and well-sampled taxa in general.