Search

Pedestrian trajectory prediction, which aims at predicting the future positions of all pedestrians in a crowd scene given their past trajectories, is the cornerstone of autonomous driving and intelligent transportation systems. Accurate prediction and fast inference are both indispensable for real-world applications. In this paper, we propose a stochastic non-autoregressive Transformer-based multi-modal trajectory prediction model to address the two challenges. Specifically, a novel graph attention module dedicated to joint learning of social and temporal interaction is proposed to explore the complex interaction among pedestrians while integrating sparse attention mechanism, pedestrian identity, and temporal order contained in the trajectory data. By doing so, the interaction across temporal and social dimensions can be simultaneously processed to extract abundant context features for prediction. Besides, to accelerate inference speed, we put forward a stochastic non-autoregressive Transformer model with multi-modal prediction capability where each future trajectory can be inferred in a parallel fashion, therefore, resulting in diverse trajectory predictions and less computational cost. Extensive experiments and ablation studies are performed to evaluate our approach. The empirical results demonstrate that the proposed model not only produces high prediction accuracy but also infers with fast speed. The code of the proposed method will be publicly available at https://github.com/xbchen82/SNARTF.

Accurately predicting the trajectory of road agents in complex traffic scenarios is challenging because the movement patterns of agents are complex and stochastic, not only depending on their own past trajectories but also being closely related to the social interaction with other surrounding agents. Besides accuracy, efficient prediction with low inference latency is also a highly desirable feature for the practical application of trajectory prediction. To address these issues, we propose the VNAGT model, a variational non-autoregressive graph transformer that adopts the framework of conditional variational autoencoder and incorporates the non-autoregressive approach such that diverse trajectories and low prediction latency can be achieved simultaneously. In order to capture the social and temporal interaction, we put forward a unified graph attention-based module that is applicable for homogeneous and heterogeneous multi-agents such that the class information can be seamlessly integrated when it is available. Non-autoregressive decoding is combined with variational learning to produce multiple plausible predictions with low latency. We train and validate the model on two real-world homogeneous and heterogeneous trajectory datasets. The experimental results demonstrate the superior performance of our model in comparison with the state-of-the-art methods.

The article focuses on the development of a three-dimensional (3-D) printer that can produce and cook dishes by Columbia University mechanical engineers in 2021. The printer uses a multiwavelength laser cooker and a microwave-oven-sized food printer. Food scientist Megan Ross describes the advantages and disadvantages of the technology. The researchers discovered that the food laser-cooked in the 3-D printer retained almost twice as much weight and volume compared with a traditional oven.

An emerging paradigm suggests that gut glycosylation is a key force in maintaining the homeostatic relationship between the gut and its microbiota. Nevertheless, it is unclear how gut glycosylation contributes to the HIV-associated microbial translocation and inflammation that persist despite viral suppression and contribute to the development of several comorbidities. We examined terminal ileum, right colon, and sigmoid colon biopsies from HIV-infected virally-suppressed individuals and found that gut glycomic patterns are associated with distinct microbial compositions and differential levels of chronic inflammation and HIV persistence. In particular, high levels of the pro-inflammatory hypo-sialylated T-antigen glycans and low levels of the anti-inflammatory fucosylated glycans were associated with higher abundance of glycan-degrading microbial species (in particular, Bacteroides vulgatus), a less diverse microbiome, higher levels of inflammation, and higher levels of ileum-associated HIV DNA. These findings are linked to the activation of the inflammasome-mediating eIF2 signaling pathway. Our study thus provides the first proof-of-concept evidence that a previously unappreciated factor, gut glycosylation, is a force that may impact the vicious cycle between HIV infection, microbial translocation, and chronic inflammation.

The gut of healthy human neonates is usually devoid of viruses at birth, but quickly becomes colonized, which—in some cases—leads to gastrointestinal disorders1–4. Here we show that the assembly of the viral community in neonates takes place in distinct steps. Fluorescent staining of virus-like particles purified from infant meconium or early stool samples shows few or no particles, but by one month of life particle numbers increase to 109per gram, and these numbers seem to persist throughout life5–7. We investigated the origin of these viral populations using shotgun metagenomic sequencing of virus-enriched preparations and whole microbial communities, followed by targeted microbiological analyses. Results indicate that, early after birth, pioneer bacteria colonize the infant gut and by one month prophages induced from these bacteria provide the predominant population of virus-like particles. By four months of life, identifiable viruses that replicate in human cells become more prominent. Multiple human viruses were more abundant in stool samples from babies who were exclusively fed on formula milk compared with those fed partially or fully on breast milk, paralleling reports that breast milk can be protective against viral infections8–10. Bacteriophage populations also differed depending on whether or not the infant was breastfed. We show that the colonization of the infant gut is stepwise, first mainly by temperate bacteriophages induced from pioneer bacteria, and later by viruses that replicate in human cells; this second phase is modulated by breastfeeding.