Bridging Global Context Interactions for High-Fidelity Pluralistic Image Completion.

We introduce PICFormer, a novel framework for Pluralistic Image Completion using a transFormer based architecture, that achieves both high quality and diversity at a much faster inference speed. Our key contribution is to introduce a code-shared codebook learning using a restrictive CNN on small and non-overlapping receptive fields (RFs) for the local visible token representation. This results in a compact yet expressive discrete representation, facilitating efficient modeling of global visible context relations by the transformer. Unlike the prevailing autoregressive approaches, we proposed to sample all tokens simultaneously, leading to more than 100× faster inference speed. To enhance appearance consistency between visible and generated regions, we further propose a novel attention-aware layer (AAL), designed to better exploit distantly related high-frequency features. Through extensive experiments, we demonstrate that the efficiently learns semantically-rich discrete codes, resulting in significantly improved image quality. Moreover, our diverse image completion framework surpasses state-of-the-art methods on multiple image completion datasets. The project page is available at https://chuanxiaz.com/picformer/.

Metagenomic insights into the modulatory effects of kelp powder (Thallus laminariae)-Treated dairy milk on growth performances and physiological lipometabolic processes of kunming mice.

Kelp powder, supplemented with a dairy cow diet, effectively improved the milk polyunsaturated fatty acids (PUFAs) content. However, little information exists on the downstream effects of the kelp-treated milk on body health, gut microbiota, and nutrient metabolism. For this purpose, 48 3-week old Kunming (KM) male mice with an average body weight of 16.1 g ± 0.2 g were randomly divided into the control treatment (CON, fed with standard chow), the common milk supplement treatment (Milk), and the kelp powder-treated milk supplement treatment (KPM). The experiment lasted for 35 days, with a 7-day long adaptive period and a 28-day long main trial. Phenotypic parameters including growth performances and serum lipids-related parameters were first measured, and results indicated that Milk and KPM supplement significantly promoted the total body weight gain (P < 0.05), while significantly decreasing the feed conversion ratio compared with CON (P < 0.05). No significant differences were observed in the blood lipids content among all three treatments, however, the triglyceride content showed a decreasing trend after KPM supplement treatment. Further, activities of liver lipometabolic-related enzymes were investigated to determine the underlying factors that impacted physiological lipid metabolism. KPM treatment showed a significant reductive effect on the activity of lipogenesis-related enzymes, such as FAS and ACC, while a significant stimulative effect on the activity of lipolysis-related enzymes included the ATGL and CPT1 compared with CON (P < 0.05). Finally, gastrointestinal tract development and cecal microbiota community that correlated with body lipid degradation and absorption were measured to determine the underlying mechanism of KPM supplementation on physiological lipid metabolism. Results indicated that supplementation with KPM significantly enhanced cecal bacteria diversity which was reflected in the significant increase of Chao1 and ACE indexes. Besides, starch-degraded bacteria such as Faecalibacterium, Ruminococcaceae, and Streptococcus are significant decreased (P < 0.05), while cellulose-degraded bacteria including Parabacteroides, Prevotella, Lactobacillus, Clostridium, and Bifidobacterium are significantly increased (P < 0.05) after KPM supplement, which may further restrict the energy generation and therefore reduce the lipid deposition. In summary, kelp supplement helped increase the milk PUFAs content, enhance the bacterial diversity and relative abundances of probiotics, which finally modulated physiological lipid metabolism, and promote growth performances.

From qualitative data to correlation using deep generative networks: Demonstrating the relation of nuclear position with the arrangement of actin filaments.

The cell nucleus is a dynamic structure that changes locales during cellular processes such as proliferation, differentiation, or migration, and its mispositioning is a hallmark of several disorders. As with most mechanobiological activities of adherent cells, the repositioning and anchoring of the nucleus are presumed to be associated with the organization of the cytoskeleton, the network of protein filaments providing structural integrity to the cells. However, demonstrating this correlation between cytoskeleton organization and nuclear position requires the parameterization of the extraordinarily intricate cytoskeletal fiber arrangements. Here, we show that this parameterization and demonstration can be achieved outside the limits of human conceptualization, using generative network and raw microscope images, relying on machine-driven interpretation and selection of parameterizable features. The developed transformer-based architecture was able to generate high-quality, completed images of more than 8,000 cells, using only information on actin filaments, predicting the presence of a nucleus and its exact localization in more than 70 per cent of instances. Our results demonstrate one of the most basic principles of mechanobiology with a remarkable level of significance. They also highlight the role of deep learning as a powerful tool in biology beyond data augmentation and analysis, capable of interpreting-unconstrained by the principles of human reasoning-complex biological systems from qualitative data.

Rumen Microbial Metabolic Responses of Dairy Cows to the Honeycomb Flavonoids Supplement Under Heat-Stress Conditions.

Flavonoids played critical roles in stabilizing microbial homoeostasis when animals suffered exoteric stresses. However, whether flavonoids attenuated heat stress of dairy cows is still not clear. Therefore, in the present article, flavonoids extracted from honeycomb were supplemented to investigate the production, digestibility, and rumen microbial metabolism responses of cows under heat stress conditions. A total of 600 multiparous dairy herds were randomly allotted into the control treatment (CON), the heat stress (HS) treatment, and the honeycomb flavonoids supplement under heat stress conditions (HF) treatment for a 30-day-long trial. Each treatment contains 4 replicates, with 50 cows in each replicate. Production performances including dry matter intake (DMI), milk production, and milk quality were measured on the basis of replicate. Furthermore, two cows of each replicate were selected for the measurement of the nutrient digestibility, the ruminal fermentable parameters including ruminal pH, volatile fatty acids, and ammonia-N, and the rumen microbial communities and metabolism. Results showed that HF effectively increased DMI, milk yield, milk fat, and ruminal acetate content (p < 0.05) compared with HS. Likewise, digestibility of NDF was promoted after HF supplement compared with HS. Furthermore, relative abundances of rumen microbial diversities especially Succiniclasticum, Pseudobutyrivibrio, Acetitomaculum, Streptococcus, and Succinivibrio, which mainly participated in energy metabolism, significantly improved after HF supplement. Metabolomic investigation showed that HF supplement significantly upregulated relative content of lipometabolic-related metabolites such as phosphatidylglycerol, phosphatidylinositol, phosphatidylserine, and phosphatidylethanolamine, while it downregulated biogenic amines. In summary, HF supplement helps proliferate microbial abundances, which further promoted fiber digestibility and energy provision, and ultimately enhances the production performances of dairy cows under heat stress conditions.