Yufei Wang

A comprehensive understanding of the evolution of the immune landscape in humans across the entire lifespan at single-cell transcriptional and protein levels, during development, maturation and senescence is currently lacking. We recruited a total of 220 healthy volunteers from the Shanghai Pudong Cohort (NCT05206643), spanning 13 age groups from 0 to over 90 years, and profiled their peripheral immune cells through single-cell RNA-sequencing coupled with single T cell and B cell receptor sequencing, high-throughput mass cytometry, bulk RNA-sequencing and flow cytometry validation experiments. We revealed that T cells were the most strongly affected by age and experienced the most intensive rewiring in cell–cell interactions during specific age. Different T cell subsets displayed different aging patterns in both transcriptomes and immune repertoires; examples included GNLY+CD8+ effector memory T cells, which exhibited the highest clonal expansion among all T cell subsets and displayed distinct functional signatures in children and the elderly; and CD8+ MAIT cells, which reached their peaks of relative abundance, clonal diversity and antibacterial capability in adolescents and then gradually tapered off. Interestingly, we identified and experimentally verified a previously unrecognized ‘cytotoxic’ B cell subset that was enriched in children. Finally, an immune age prediction model was developed based on lifecycle-wide single-cell data that can evaluate the immune status of healthy individuals and identify those with disturbed immune functions. Our work provides both valuable insights and resources for further understanding the aging of the immune system across the whole human lifespan. In this Resource, authors profile peripheral immune cells from a total of 220 healthy volunteers from birth to over 90 years. This revealed that T cells were most affected by aging with divergent aging patterns in different subsets and identified a population of cytotoxic B cells that were enriched in children.

RGB-guided depth completion aims at predicting dense depth maps from sparse depth measurements and corresponding RGB images, where how to effectively and efficiently exploit the multi-modal information is a key issue. Guided dynamic filters, which generate spatially-variant depth-wise separable convolutional filters from RGB features to guide depth features, have been proven to be effective in this task. However, the dynamically generated filters require massive model parameters, computational costs and memory footprints when the number of feature channels is large. In this paper, we propose to decompose the guided dynamic filters into a spatially-shared component multiplied by content-adaptive adaptors at each spatial location. Based on the proposed idea, we introduce two decomposition schemes <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mathcal {A}$ </tex-math></inline-formula> and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mathcal {B}$ </tex-math></inline-formula> , which decompose the filters by splitting the filter structure and using spatial-wise attention, respectively. The decomposed filters not only maintain the favorable properties of guided dynamic filters as being content-dependent and spatially-variant, but also reduce model parameters and hardware costs, as the learned adaptors are decoupled with the number of feature channels. Extensive experimental results demonstrate that the methods using our schemes outperform state-of-the-art methods on the KITTI dataset, and rank 1st and 2nd on the KITTI benchmark at the time of submission. Meanwhile, they also achieve comparable performance on the NYUv2 dataset. In addition, our proposed methods are general and could be employed as plug-and-play feature fusion blocks in other multi-modal fusion tasks such as RGB-D salient object detection.

We address the problem of goal-directed cloth manipulation, a challenging task due to the deformability of cloth. Our insight is that optical flow, a technique normally used for motion estimation in video, can also provide an effective representation for corresponding cloth poses across observation and goal images. We introduce FabricFlowNet (FFN), a cloth manipulation policy that leverages flow as both an input and as an action representation to improve performance. FabricFlowNet also elegantly switches between bimanual and single-arm actions based on the desired goal. We show that FabricFlowNet significantly outperforms state-of-the-art model-free and model-based cloth manipulation policies that take image input. We also present real-world experiments on a bimanual system, demonstrating effective sim-to-real transfer. Finally, we show that our method generalizes when trained on a single square cloth to other cloth shapes, such as T-shirts and rectangular cloths. Video and other supplementary materials are available at: https://sites.google.com/view/fabricflownet.

Robotic manipulation of cloth remains challenging for robotics due to the complex dynamics of the cloth, lack of a low-dimensional state representation, and self-occlusions. In contrast to previous model-based approaches that learn a pixel-based dynamics model or a compressed latent vector dynamics, we propose to learn a particle-based dynamics model from a partial point cloud observation. To overcome the challenges of partial observability, we infer which visible points are connected on the underlying cloth mesh. We then learn a dynamics model over this visible connectivity graph. Compared to previous learning-based approaches, our model poses strong inductive bias with its particle based representation for learning the underlying cloth physics; it is invariant to visual features; and the predictions can be more easily visualized. We show that our method greatly outperforms previous state-of-the-art model-based and model-free reinforcement learning methods in simulation. Furthermore, we demonstrate zero-shot sim-to-real transfer where we deploy the model trained in simulation on a Franka arm and show that the model can successfully smooth different types of cloth from crumpled configurations. Videos can be found on our project website.