Generative adversarial networks conditioned on textual image descriptions are capable of generating realistic-looking images. However, current methods still struggle to generate images based on complex image captions from a heterogeneous domain. Furthermore, quantitatively evaluating these text-to-image models is challenging, as most evaluation metrics only judge image quality but not the conformity between the image and its caption. To address these challenges we introduce a new model that explicitly models individual objects within an image and a new evaluation metric called Semantic Object Accuracy (SOA) that specifically evaluates images given an image caption. The SOA uses a pre-trained object detector to evaluate if a generated image contains objects that are mentioned in the image caption, e.g. whether an image generated from "a car driving down the street" contains a car. We perform a user study comparing several text-to-image models and show that our SOA metric ranks the models the same way as humans, whereas other metrics such as the Inception Score do not. Our evaluation also shows that models which explicitly model objects outperform models which only model global image characteristics.

Kelps, brown algae of the order Laminariales, dominate rocky shores and form huge kelp beds which provide habitat and nurseries for various marine organisms. Whereas the basic physiological and ecophysiological characteristics of kelps are well studied, the molecular processes underlying acclimation to different light and temperature conditions are still poorly understood. Therefore we investigated the molecular mechanisms underlying the physiological acclimation to light and temperature stress. Sporophytes of S. latissima were exposed to combinations of light intensities and temperatures, and microarray hybridizations were performed to determine changes in gene expression patterns. This first large-scale transcriptomic study of a kelp species shows that S. latissima responds to temperature and light stress with a multitude of transcriptional changes: up to 32% of genes showed an altered expression after the exposure experiments. High temperature had stronger effects on gene expression in S. latissima than low temperature, reflected by the higher number of temperature-responsive genes. We gained insights into underlying molecular processes of acclimation, which includes adjustment of the primary metabolism as well as induction of several ROS scavengers and a sophisticated regulation of Hsps. We show that S. latissima, as a cold adapted species, must make stronger efforts for acclimating to high than to low temperatures. The strongest response was caused by the combination of high temperatures with high light intensities, which proved most harmful for the alga.

No abstract

The Arctic region is currently facing substantial environmental changes due to global warming. Melting glaciers cause reduced salinity environments in coastal Arctic habitats, which may be stressful for kelp beds. To investigate the responses of the kelp Saccharina latissima to the warming Arctic, we studied the transcriptomic changes of S. latissima from Kongsfjorden (Svalbard, Norway) over a 24‐hour exposure to two salinities (Absolute Salinity [SA] 20 and 30) after a 7‐day pre‐acclimation at three temperatures (0, 8 and 15°C). In addition, corresponding physiological data were assessed during an 11‐days salinity/temperature experiment. Growth and maximal quantum yield for photosystem II fluorescence were positively affected by increased temperature during acclimation, whereas hyposalinity caused negative effects at the last day of treatment. In contrast, hyposalinity induced marked changes on the transcriptomic level. Compared to the control (8°C – SA 30), the 8°C – SA 20 exhibited the highest number of differentially expressed genes (DEGs), followed by the 0°C – SA 20. Comparisons indicate that S. latissima tends to convert its energy from primary metabolism (e.g. photosynthesis) to antioxidant activity under hyposaline stress. The increase in physiological performance at 15°C shows that S. latissima in the Arctic region can adjust and might even benefit from increased temperatures. However, in Arctic fjord environments its performance might become impaired by decreased salinity as a result of ice melting.

The delta(13)C(VPDB), delta(2)H(VSMOW) and delta(18)O(VSMOW) values of caffeine isolated from Arabica green coffee beans of different geographical origin have been determined by isotope ratio mass spectrometry (IRMS) using elemental analysis (EA) in the "combustion" (C) and "pyrolysis" (P) modes (EA-C/P-IRMS). In total, 45 coffee samples (20 from Central and South America, 16 from Africa, six from Indonesia, and three from Jamaica and Hawaii) were analysed, as well as three reference samples of synthetic caffeine. Validation was performed by excluding isotope discrimination in the course of sample preparation and determining linear dynamic ranges for EA-P-IRMS measurements. The values for caffeine from green coffee ranged from -25.1 to - 29.9 per thousand, -109 to -198 per thousand, and +2.0 to -12.0 per thousand for delta(13)C(VPDB), delta(2)H(VSMOW), and delta(18)O(VSMOW), respectively. Data evaluation by linear discrimination analysis (LDA) and by classification and regression tree (CART) analysis revealed the delta(18)O(VSMOW) values to be highly significant. Use of LDA on the delta(2)H(VSMOW) and delta(18)O(VSMOW) data from coffee of African and Central/South American provenance led to error rates of 5.7% and 7.7% for adaption and cross validation, respectively.