Color constancy is a well-studied topic in color vision. Methods are generally categorized as (1) low-level statistical methods, (2) gamut-based methods, and (3) learning-based methods. In this work, we distinguish methods depending on whether they work directly from color values (i.e., color domain) or from values obtained from the image's spatial information (e.g., image gradients/frequencies). We show that spatial information does not provide any additional information that cannot be obtained directly from the color distribution and that the indirect aim of spatial-domain methods is to obtain large color differences for estimating the illumination direction. This finding allows us to develop a simple and efficient illumination estimation method that chooses bright and dark pixels using a projection distance in the color distribution and then applies principal component analysis to estimate the illumination direction. Our method gives state-of-the-art results on existing public color constancy datasets as well as on our newly collected dataset (NUS dataset) containing 1736 images from eight different high-end consumer cameras.

Variation in terrestrial net primary production (NPP) with climate is thought to originate from a direct influence of temperature and precipitation on plant metabolism. However, variation in NPP may also result from an indirect influence of climate by means of plant age, stand biomass, growing season length and local adaptation. To identify the relative importance of direct and indirect climate effects, we extend metabolic scaling theory to link hypothesized climate influences with NPP, and assess hypothesized relationships using a global compilation of ecosystem woody plant biomass and production data. Notably, age and biomass explained most of the variation in production whereas temperature and precipitation explained almost none, suggesting that climate indirectly (not directly) influences production. Furthermore, our theory shows that variation in NPP is characterized by a common scaling relationship, suggesting that global change models can incorporate the mechanisms governing this relationship to improve predictions of future ecosystem function.

Background: Accurate glioma grading before surgery is of the utmost importance in treatment planning and prognosis prediction. But previous studies on magnetic resonance imaging (MRI) images were not effective enough. According to the remarkable performance of convolutional neural network (CNN) in medical domain, we hypothesized that a deep learning algorithm can achieve high accuracy in distinguishing the World Health Organization (WHO) low grade and high grade gliomas.Methods: One hundred and thirteen glioma patients were retrospectively included. Tumor images were segmented with a rectangular region of interest (ROI), which contained about 80% of the tumor. Then, 20% data were randomly selected and leaved out at patient-level as test dataset. AlexNet and GoogLeNet were both trained from scratch and fine-tuned from models that pre-trained on the large scale natural image database, ImageNet, to magnetic resonance images. The classification task was evaluated with five-fold cross-validation (CV) on patient-level split.Results: The performance measures, including validation accuracy, test accuracy and test area under curve (AUC), averaged from five-fold CV of GoogLeNet which trained from scratch were 0.867, 0.909, and 0.939, respectively. With transfer learning and fine-tuning, better performances were obtained for both AlexNet and GoogLeNet, especially for AlexNet. Meanwhile, GoogLeNet performed better than AlexNet no matter trained from scratch or learned from pre-trained model.Conclusion: In conclusion, we demonstrated that the application of CNN, especially trained with transfer learning and fine-tuning, to preoperative glioma grading improves the performance, compared with either the performance of traditional machine learning method based on hand-crafted features, or even the CNNs trained from scratch.

The allometries of forest- and individual plant-level M(A) vs. M(R) relationships share strikingly similar scaling exponents, but differ because of considerable variation in y-intercepts. These results support prior allometric theory and provide boundary conditions for the scaling of M(A) and M(R).

No abstract

These authors contributed equally to the study.The WBE theory proposed by West, Brown and Enquist predicts that larger plant respiration rate, R, scales to the three-quarters power of body size, M. However, studies on the R versus M relationship for larger plants (i.e. trees larger than saplings) have not been reported. Published respiration rates of field-grown trees (saplings and larger trees) were examined to test this relationship. Our results showed that for larger trees, aboveground respiration rates R A scaled as the 0.82-power of aboveground biomass M A , and that total respiration rates R T scaled as the 0.85-power of total biomass M T , both of which significantly deviated from the three-quarters scaling law predicted by the WBE theory, and which agreed with 0.81-0.84-power scaling of biomass to respiration across the full range of measured tree sizes for an independent dataset reported by Reich et al. (Reich et al. 2006 Nature 439, 457-461). By contrast, R scaled nearly isometrically with M in saplings. We contend that the scaling exponent of plant metabolism is close to unity for saplings and decreases (but is significantly larger than three-quarters) as trees grow, implying that there is no universal metabolic scaling in plants.

No abstract

The data collected for all five bamboo species are consistent with the "diminishing returns" hypothesis, i.e., the scaling exponents governing the leaf area vs. dry mass scaling relationship are less than one within and across species and are insensitive to light conditions or elevation.